JP7499210B2 - Crack detection method, crack detection device, and program - Google Patents

Description

The present invention relates to a crack detection method, a crack detection device, and a program.

Currently, various infrastructure facilities such as roads, bridges, and tunnels are aging, and repair works are being carried out in various places, and repair plans are being promoted. Most of the infrastructure facilities are reinforced concrete structures (RC (Reinforced Concrete) structures) and steel structures, and various damaged parts are present on the surface of RC structures that have been completed for more than 40 years. Specific examples of such damaged parts include cracks, efflorescence, free lime, rust, etc. on the concrete surface.
When carrying out repair work or planning repairs to RC structures, etc., the surface of the RC structure is first inspected by technical personnel of the infrastructure facility or by technical personnel of the outsourced survey company or construction company. In this inspection, the width and length of cracks, the shape and area of free lime, etc. are quantitatively evaluated, and based on this quantitative evaluation, a decision is made as to whether the structure needs repair work or maintenance.

A variety of methods have been proposed to quantitatively detect cracks on the concrete surface of RC structures. All of these crack detection methods share the common process of performing wavelet transformation processing and binarization processing, and then perform noise removal processing as necessary to detect cracks.

JP 2006-162583 A JP 2008-267943 A JP 2008-185510 A JP 2010-121992 A JP 2012-002531 A JP 2013-117409 A JP 2013-002839 A

The crack detection methods described in Patent Documents 1 to 7 make it possible to detect cracks with high accuracy even when cracks are difficult to detect due to dirt on the concrete surface or lighting conditions.

In the crack image analysis used to detect the above cracks, the crack width is generally detected in pixel units (picture element units). Therefore, cracks whose end points are physically separated from each other are not determined to be continuous cracks in the image analysis, and are generally output as different cracks when output.
However, even in the case of cracks whose end points are completely separated like this, when an inspector actually inspects, there are often cases where the cracks are determined to be essentially continuous based on the linearity (vector) and widths of both cracks. When an inspector actually inspects, cracks that can be determined to be essentially continuous like this are treated as cracks with the same label in a crack distribution diagram, and the crack width (for example, the maximum crack width) is identified.
In this way, image analysis has improved the accuracy of crack detection and significantly reduced the time it takes to detect cracks. However, there is a demand for a crack detection method that can output cracks with the same label if the end points of the cracks are physically separated but are related enough to be determined to be a continuous crack.

The present invention aims to provide a crack detection method, crack detection device, and program that can detect and output cracks with high accuracy on a pixel-by-pixel basis, as well as detect and output cracks with the same label that can be determined to be essentially continuous cracks, even if the end points of the cracks are physically separated.

In order to achieve the above object, one aspect of the crack detection method according to the present invention includes:
an input image creation step of inputting a photographed image of a concrete surface containing cracks into a computer to create an input image;
a wavelet image creation process for performing a wavelet transform process on the input image, or for extracting assumed cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on a concrete surface, drawing along the assumed cracks to create assumed crack drawing lines, and for performing a wavelet transform process on a drawing-containing image including the assumed crack drawing lines, or for performing a wavelet transform process on a path image including paths added along assumed crack lines that are assumed to be cracks to the input image, to create a wavelet image in which each pixel has a wavelet coefficient;
a binary image creation step of binarizing each pixel of the wavelet image by using a wavelet coefficient table that is a threshold value of the wavelet coefficient to create a binary image;
A crack image creation process for labeling each crack with a block-like figure having geometric connectivity in the binarized image and creating a crack image including a plurality of labeled cracks;
A segmentation image creation process is provided for creating a segmentation image by connecting end points of cracks that are determined to be essentially continuous cracks among the multiple separate cracks in the crack image,
The segmentation image creation step includes:
An end-to-end distance determination step of determining an end-to-end distance between end points of the cracks that are separated from each other;
a determination step of comparing a threshold value for the distance between the end points with the distance between the end points, and determining that the two cracks whose distance between the end points is specified are essentially continuous cracks and are cracks with the same label when the distance between the end points is equal to or less than the threshold;
and a drawing process for creating the segmentation image by connecting the end points of two cracks that have been determined to be cracks of the same label.

According to this embodiment, a wavelet image is created from an input image, a binary image is created from the wavelet image, and a crack image is created from the binary image.After that, a segmentation image is created by connecting the end points of multiple separated (spaced) cracks in the crack image that are determined to be essentially continuous cracks.This makes it possible to not only detect and output cracks on a pixel-by-pixel basis, but also to detect and output cracks with the same label, even if the end points of the cracks are physically separated, but which can be determined to be essentially continuous cracks.
In the segmentation image creation process for creating a segmentation image, the distance between the endpoints of the crack is identified in the endpoint distance identification process, and the endpoint distance is compared with a threshold value for the endpoint distance in the judgment process to determine whether the two cracks whose endpoint distances have been identified are cracks of the same label.
The above image analysis makes it possible to accurately identify and output cracks whose endpoints are separated as separate cracks on a pixel-by-pixel basis. In addition, in the segmentation image, cracks that meet the conditions can be identified and output as cracks with the same label rather than on a pixel-by-pixel basis, similar to the results of an inspection by an inspector.
It is preferable that, when outputting, it is possible to selectively output either one or both of a pixel-by-pixel output method and a segmentation image output method.

In addition, in the wavelet image creation process, the assumed cracks that are assumed to be cracks are extracted from the input image, an assumed crack drawing line is created by drawing along the assumed crack, and a wavelet transform process is performed on the drawing-containing image including this assumed crack drawing line, thereby narrowing the crack image analysis range (or analysis target) as much as possible. Therefore, it is possible to significantly shorten the analysis time compared to conventional methods. In addition, when drawing along the assumed cracks to create the assumed crack drawing line, instead of the analyst drawing, a computer that has learned the characteristics of cracks on the concrete surface draws based on this learning information, eliminating the need for the analyst to draw, and making it possible to create the assumed crack drawing line in a much shorter time than drawing by the analyst. Furthermore, by performing a wavelet transform process on the drawing-containing image created in this way to create a binary image, cracks can be detected with high accuracy.

In addition, when creating a path image for an input image in the wavelet image creation process, a path is created along the expected crack line, and a surface is created to cover the crack on the path, thereby allowing the crack position to be specified in the input image. Specifically, a path consisting of, for example, a line or surface of one pixel is created using commercially available image editing software. A path image is, for example, a surface consisting of a one-pixel-wide line, with one pixel or two pixels added to the left and right sides of the path, making it three or five pixels wide. In this way, by specifying the crack position in the input image and then performing wavelet transformation to create a wavelet image, the crack image analysis range can be reduced as much as possible, and the analysis area can be significantly reduced, which makes it possible to significantly reduce the analysis time, and furthermore, makes it possible to perform continuous analysis, thereby reducing the data reading and writing times.

Another aspect of the crack detection method according to the present invention is to
In the determination step,
If the distance between the endpoints is less than or equal to the threshold value, an endless contour range is set starting from an endpoint of one of the cracks, and if the endpoint of the other crack falls within the endless contour range, the two cracks whose endpoint distance has been identified are determined to be cracks of the same label.

According to this aspect, in order to determine whether mutually separated cracks are cracks of the same label (whether the cracks have continuity), in addition to comparing the distance between the endpoints with a threshold, an endless contour range is set starting from the endpoint of one crack, and if the endpoint of the other crack falls within this contour range, both cracks are determined to be cracks of the same label, making it possible to segment multiple cracks with even greater accuracy.
This method considers the extension direction (vector) of the crack near the end points of both cracks to verify the presence or absence of continuity with distant cracks. Even if the distance between the end points is below the threshold, if the vectors near the end points of both cracks are in completely different directions (directions that do not intersect at all), the possibility that the cracks are continuous is extremely low, and even when an inspector inspects, such cracks will not be judged as cracks with the same label, so the accuracy of segmentation is improved by taking into account the vectors near the end points of the cracks.
Here, a method for setting a vector near the end of each crack may include determining a vector by connecting an end point of the crack to a point several pixels inside the crack from the end point.

In another aspect of the crack detection method according to the present invention,
The endless contour range is characterized in that a regular polygon with a side of a predetermined length or a circle with a diameter of a predetermined length is set in the extension direction of the crack, starting from an end point of one of the cracks, and the contour range is a regular polygon or circle.

According to this aspect, by setting a regular polygon with a side of a predetermined length or a circle with a diameter of a predetermined length as an endless contour range, which can define a certain contour range, the endless contour range can be easily created, and the presence or absence of continuity with other cracks can be determined by giving a certain width (angle) to the vector near the end point of the crack. Here, as the regular polygon, it is preferable to apply an equilateral triangle or a square, which are easier to create. When applying an equilateral triangle, for example, an equilateral triangle can be set by positioning the vertex of the equilateral triangle at the end point of the crack and positioning the center of the base of the equilateral triangle on a virtual vector extended from the vicinity of the end point. Furthermore, when applying a square, for example, a square can be set by positioning one vertex of the square at the end point of the crack and positioning another vertex at the diagonal position of this vertex on a virtual vector extended from the vicinity of the end point. Furthermore, when applying a circle, a circle can be set by positioning one point on the circumference at the end point of the crack and positioning the diameter passing through this one point on a virtual vector extended from the vicinity of the end point.

Another aspect of the crack detection method according to the present invention is to
An output step is further included.
When the segmentation image is created, either or both of the crack image and the segmentation image are selectively output.

According to this embodiment, in addition to outputting a crack image in pixel units, it is possible to output a segmentation image that includes cracks with the same label that can be determined to be essentially continuous cracks, even if the end points of the cracks are physically separated. In this case, the analyst can select to output either the crack image or the segmentation image, or both, to the computer screen. In addition, by adding the crack width (maximum crack width) of each crack in a crack distribution map, which is one example of the output, the distribution of each crack on the concrete surface and the damage condition of each crack can be confirmed on a single screen.

In addition, one aspect of the crack detection device according to the present invention is
an input image creation unit that creates an input image based on a photographed image of a concrete surface containing cracks that is input to a computer;
a wavelet image creation unit that performs a wavelet transform process on the input image, or extracts assumed cracks that are assumed to be cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on a concrete surface, draws along the assumed cracks to create assumed crack drawing lines, and performs a wavelet transform process on a drawing-containing image that includes the assumed crack drawing lines, or performs a wavelet transform process on a path image that includes paths added along assumed crack lines that are assumed to be cracks on the input image, and creates a wavelet image in which each pixel has a wavelet coefficient;
a binary image creation unit that uses a wavelet coefficient table that is a threshold value of wavelet coefficients to binarize each pixel of the wavelet image to create a binary image;
A crack image creation unit that labels each crack with a block-like figure having geometric connectivity in the binarized image and creates a crack image including the labeled cracks;
A segmentation image creation unit that creates a segmentation image by connecting end points of cracks that are determined to be essentially continuous cracks among the multiple separate cracks in the crack image,
The segmentation image creation unit includes:
An end point distance specification unit that specifies an end point distance between end points of the cracks that are separated from each other;
The present invention is characterized in that it has a judgment unit that compares a threshold value for the distance between the end points with the distance between the end points, and if the distance between the end points is equal to or less than the threshold value, judges that the two cracks whose distance between the end points is identified are essentially continuous cracks and are cracks with the same label.

According to this aspect, a wavelet image is created from an input image, a binary image is created from the wavelet image, and a crack image is created from the binary image. Then, among multiple separated cracks in the crack image, the end points of cracks that are determined to be essentially continuous are connected to create a segmentation image. In addition to detecting and outputting cracks on a pixel-by-pixel basis, it is possible to detect and output cracks that are determined to be essentially continuous even if their end points are physically separated as cracks of the same label. Here, in the segmentation image creation unit, when the determination unit determines that two mutually separated cracks are cracks of the same label, the analyst may use image editing software to connect the two cracks on the screen, or the segmentation image creation unit may further include a drawing unit, and the drawing unit may automatically perform a process of connecting the two cracks.

Another aspect of the crack detection device according to the present invention is
In the determination unit,
If the distance between the endpoints is less than or equal to the threshold value, an endless contour range is set starting from an endpoint of one of the cracks, and if the endpoint of the other crack falls within the endless contour range, the two cracks whose endpoint distance has been identified are determined to be cracks of the same label.

According to this aspect, in order to determine whether mutually separated cracks are cracks of the same label, in addition to comparing the distance between the endpoints with a threshold, an endless contour range is set starting from the endpoint of one crack, and if the endpoint of the other crack falls within this contour range, both cracks are determined to be cracks of the same label, making it possible to segment multiple cracks with even greater accuracy.

Another aspect of the crack detection device according to the present invention is
Further comprising an output section,
When the segmentation image is created, either or both of the crack image and the segmentation image are selectively output.

According to this embodiment, in addition to outputting a crack image on a pixel-by-pixel basis, it is possible to output a segmentation image that includes cracks with the same label that can be determined to be essentially continuous cracks, even if the end points of the cracks are physically separated. In this case, depending on the analyst's choice, it is possible to output either the crack image or the segmentation image, or both, to the computer screen.

Moreover, one aspect of the program according to the present invention is
A program for causing a computer to execute the following processes for detecting cracks on a concrete surface,
an input image creation step of inputting a photographed image of a concrete surface containing cracks into a computer to create an input image;
a wavelet image creation process for performing a wavelet transform process on the input image, or for extracting assumed cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on a concrete surface, drawing along the assumed cracks to create assumed crack drawing lines, and for performing a wavelet transform process on a drawing-containing image including the assumed crack drawing lines, or for performing a wavelet transform process on a path image including paths added along assumed crack lines that are assumed to be cracks to the input image, to create a wavelet image in which each pixel has a wavelet coefficient;
a binary image creation step of binarizing each pixel of the wavelet image by using a wavelet coefficient table that is a threshold value of the wavelet coefficient to create a binary image;
A crack image creation process for labeling each crack with a block-like figure having geometric connectivity in the binarized image and creating a crack image including a plurality of labeled cracks;
A segmentation image creation process is provided for creating a segmentation image by connecting end points of cracks that are determined to be essentially continuous cracks among the multiple separate cracks in the crack image,
The segmentation image creation step includes:
An end-to-end distance determination step of determining an end-to-end distance between end points of the cracks that are separated from each other;
The method is characterized by having a judgment process of comparing the distance between the end points with a threshold value for the distance between the end points, and if the distance between the end points is equal to or less than the threshold value, judging that the two cracks whose distance between the end points is identified are essentially continuous cracks and are cracks with the same label.

According to this aspect, by having a computer execute each step of the program, cracks whose endpoints are far apart can be accurately identified and output as separate cracks on a pixel-by-pixel basis. In addition, in the segmentation image, cracks that meet the conditions can be identified and output as cracks with the same label, not on a pixel-by-pixel basis, in the same way as the results of an inspection by an inspector.

Another aspect of the program according to the present invention is
In the determination step,
If the distance between the endpoints is less than or equal to the threshold value, an endless contour range is set starting from an endpoint of one of the cracks, and if the endpoint of the other crack falls within the endless contour range, the two cracks whose endpoint distance has been identified are determined to be cracks of the same label.

According to this aspect, in order to determine whether or not mutually separated cracks are cracks of the same label, in addition to comparing the distance between the endpoints with a threshold value, an endless contour range is set starting from the endpoint of one crack, and if the endpoint of the other crack falls within this contour range, both cracks are determined to be cracks of the same label, making it possible to segment multiple cracks with even greater accuracy.

As can be understood from the above explanation, the crack detection method, crack detection device, and program of the present invention can detect and output cracks with high accuracy on a pixel-by-pixel basis, and can also detect and output cracks that are essentially continuous, even if their endpoints are physically separated, as cracks with the same label.

A diagram showing an example of the overall configuration of a crack detection system including a crack detection device according to an embodiment. FIG. 2 is a diagram illustrating an example of a hardware configuration of a crack detection device. FIG. 2 is a diagram illustrating an example of a functional configuration of a crack detection device. 4 is a flowchart showing an example of a crack detection method according to the embodiment. 1A and 1B are diagrams illustrating an example of a relationship between an input image and a local region in wavelet transform processing. 1A and 1B are diagrams illustrating an example of a relationship between a local region and a pixel of interest in wavelet transform processing. FIG. 13 is a diagram showing an example of a pseudo image. FIG. 7 is a diagram showing an example of a bird's-eye view of wavelet coefficients of the pseudo image of FIG. 6 . FIG. 11 is a diagram illustrating an example of a wavelet coefficient table. (a) is a schematic diagram showing an example in which cracks that are separated from each other are determined to be essentially continuous cracks, and (b) and (c) are both schematic diagrams showing examples in which cracks that are separated from each other are determined not to be essentially continuous cracks. 9A, 9B, and 9C are schematic diagrams illustrating an example of a determination that is further performed in the determination unit when the distance between the end points is equal to or smaller than a threshold value, using FIG. 9A. 9A and 9B are schematic diagrams illustrating an example of a determination that is further performed in the determination unit when the distance between the end points is equal to or smaller than a threshold value, using FIGS. 9B and 9C, respectively. 10 is a photograph showing a pixel-by-pixel crack image output by the output unit. FIG. FIG. 13 is a photograph showing a segmentation image output by an output unit. This is a photograph showing the concrete surface of test specimen 1 used in a verification experiment to verify the accuracy of the crack width identified by the segmentation image. This is a photograph showing the concrete surface of test specimen 2 used in a verification experiment to verify the accuracy of the crack width identified by the segmentation image. This is a photograph showing an example output in which the maximum crack width is noted after deleting the top 10% of cracks with the largest crack widths in the segmentation image of test specimen 1. This is a photograph showing an example output in which the maximum crack width is noted after deleting the top 10% of cracks with the largest crack widths in the segmentation image of test specimen 2.

Below, the crack detection method, crack detection device, and program according to the embodiment will be described with reference to the attached drawings. Note that in this specification and the drawings, substantially identical components may be denoted by the same reference numerals to avoid redundant description.

[Embodiment]
<Crack detection system>
First, a crack detection system including a crack detection device according to an embodiment will be described. FIG. 1 is a diagram showing an example of the overall configuration of a crack detection system including a crack detection device according to an embodiment. As shown in FIG. 1, the crack detection system 1000 has an imaging device 100 and a crack detection device 300. The crack detection device 300 is, for example, a server device, and the imaging device 100 and the crack detection device 300 are connected via a network 200 such as a public network such as the Internet, a wireless network such as a mobile phone network, a dedicated network such as a VPN (Virtual Private Network), or a LAN (Local Area Network).

The imaging device 100 has an imaging section such as a CCD camera, digital camera, or digital video camera, and a communication section that transmits image data captured by the imaging section. The imaging device 100 may have only a shooting section, and may be connected to a mobile terminal or the like, and image data may be transmitted from the mobile terminal or the like. The image data captured by the imaging device 100 is transmitted to the crack detection device 300 via the network 200. This image data includes image information of the concrete surface. For example, the concrete surface of infrastructure facilities such as road bridges and tunnels made of reinforced concrete that were built decades ago is imaged by the imaging device 100 to determine whether or not they require repair work.

A data collection program and a data analysis program are installed in the crack detection device 300, and the crack detection device 300 functions as a data collection unit 301 and a data processing unit 302 by executing these programs.

The data collection unit 301 receives image data captured by the imaging device 100 and transmitted from the imaging device 100 or a mobile terminal or the like connected to the imaging device 100, and stores the image data in the data storage unit 303. The data processing unit 302 also executes a process of sequentially creating various images based on the image data stored in the data storage unit 303, and outputting the various images that have been created.

<Crack detection device>
Next, an example of a crack detection device according to an embodiment will be described with reference to Figures 2 and 3 .

(Hardware configuration of crack detection device)
First, an example of the hardware configuration of the crack detection device will be described with reference to Fig. 2. As shown in Fig. 2, the crack detection device 300 has a CPU (Central Processing Unit) 401, a ROM (Read Only Memory) 402, a RAM (Random Access Memory) 403, an auxiliary storage unit 404, a display unit 405, and a communication unit 406, and each unit is connected to each other via a bus 407.

The CPU 401 executes the various programs installed in the auxiliary storage unit 404. The ROM 402 is a non-volatile memory and functions as a main storage unit that stores various programs and data necessary for the CPU 401 to execute the various programs stored in the auxiliary storage unit 404. The RAM 403 is a volatile memory and functions as a main storage unit. The RAM 403 functions as a working area when the various programs stored in the auxiliary storage unit 404 are executed by the CPU 401. The auxiliary storage unit 404 stores the various programs installed in the crack detection device 300 and data used when executing the various programs.

The display unit 405 displays various screens. For example, it displays image data transmitted from the imaging device 100 as a captured image, and also displays an input image, a wavelet image, a binarized image, a crack image, a segmentation image, etc.

The communication unit 406 is connected to the imaging device 100 or a mobile terminal connected to the imaging device 100, and receives image data from the imaging device 100, and transmits the length of each crack width and the total crack length identified and created by the crack detection device 300 to the mobile terminal connected to the imaging device 100.

(Functional configuration of the crack detection device)
Next, an example of the functional configuration of the crack detection device will be described with reference to Fig. 3. As shown in Fig. 3, image data transmitted from the imaging device 100 is received by a data collection unit 301, and temporarily stored in a data storage unit 303 from the data collection unit 301. When an analysis is performed by the data processing unit 302, the image data stored in the data storage unit 303 is input to an input unit 502 in the data processing unit 302.

The data processing unit 302 has an input unit 502, an input image creation unit 504, a wavelet image creation unit 506, a wavelet coefficient table creation unit 508, a binary image creation unit 510, a crack image creation unit 512, a segmentation image creation unit 514, an output unit 516, a machine learning unit 530, and a drawing unit 532. The segmentation image creation unit 514 has an end point distance determination unit 520, a judgment unit 522, and a drawing unit 524.

The input unit 502 takes in a photographed image based on image data stored in the data storage unit 303. Image data of various concrete surfaces is stored in the data storage unit 303, and the analyst selects a photographed image based on the image data to be analyzed, which is then taken and input into the input unit 502.

The input image creation unit 504 creates an input image by correcting the luminance of the captured image input to the input unit 502 as necessary. For example, the luminance has 256 gradations from luminance value 0 to 255, but a predetermined luminance correction value is input in advance to the input image creation unit 504 by the analyst. For example, if 150 is input as the luminance correction value to the input image creation unit 504, the input image creation unit 504 identifies the luminance of the captured image imported from the input unit 502, performs luminance correction processing so that the luminance of the captured image becomes the luminance correction value of 150, and creates an input image. Note that whether or not such luminance correction is performed is optional.

Here, based on the learning information in the machine learning unit 530, an assumed crack that is assumed to be a crack in the input image may be extracted, and an assumed crack drawing line may be created by drawing along the assumed crack, and a drawing-containing image including this assumed crack drawing line may be created by the drawing unit 532.

The drawing unit 532 extracts assumed cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on the surface of concrete, draws along the assumed cracks to create assumed crack drawing lines, and creates a drawing-containing image that includes the assumed crack drawing lines.

Here, the learning information in the machine learning unit 530 will be outlined. First, a feature database relating to various concrete surface images is stored in the data storage unit 303. Crack lines that the analyst determines to be cracks are drawn on the concrete surface images. Therefore, the stored concrete surface images contain the photographed image and the drawn crack lines. The machine learning unit 530 creates a function that has machine-learned the crack features in each concrete surface image and a model that shows the crack features, based on the concrete surface images input to the data storage unit 303 and the crack lines drawn in the concrete surface images.

The function that has learned the crack characteristics through machine learning and the model that shows the crack characteristics can be created using existing technology, such as supervised learning using a neural network, but the creation method is not particularly limited. By repeating machine learning in the machine learning unit 530, various crack patterns in various concrete surface images are learned. The learning data regarding the crack patterns learned for the concrete surface images is stored again in the data storage unit 303.

The computer learns the characteristics of cracks on the surface of concrete, and then performs wavelet transformation on the image containing the drawing based on this learning information to create a binary image and a crack image, making it possible to detect cracks with high accuracy.

Alternatively, a path image may be created for the input image, including a path that is added along an assumed crack line that is assumed to be a crack. The crack position can be specified in the input image by creating a path along the assumed crack line and creating a surface that covers the crack added to the path. Specifically, a path consisting of, for example, a line or surface of one pixel is created using commercially available image editing software. Then, the path image is, for example, a path consisting of a line of one pixel in width, with a surface of three or five pixels in width, which is one pixel or two pixels wide on both sides of the path, added to the left and right of the line, to cover the crack added to the path.

By creating a rough path (a path that is several pixels wider than the width of the expected crack line) along the expected crack line to specify the crack position, problems such as the expected crack line going beyond the path and not being detected as a crack are eliminated, and because the path is rough, it can be created efficiently and does not take a long time to create the path (identify the crack position). Furthermore, by creating a rough path for the expected crack line and performing a wavelet transform and then thinning process, it becomes possible to identify the crack with even greater precision.

The wavelet image creation unit 506 creates a wavelet image by performing a wavelet transform process on the input image created by the input image creation unit 504. A wavelet means a small wave, and a basic unit of a localized wave can be expressed by an equation using a wavelet function. By expanding or contracting this wavelet function, it becomes possible to simultaneously analyze time information, spatial information, and frequency information. When this wavelet coefficient is applied to a concrete surface having cracks, the coefficient is characterized in that it depends on the density of the concrete surface, the density of the cracks, and the width (or extent) of the cracks. For example, the value of the wavelet coefficient tends to increase as the width of the crack increases, and the value of the wavelet coefficient tends to increase as the density of the cracks increases (as they become darker). The algorithm for creating a binary image using the wavelet coefficient calculated by the wavelet transform process will be described in detail below.

The wavelet coefficient table creation unit 508 creates wavelet coefficients that serve as thresholds when creating a binary image from a wavelet image. Since the wavelet coefficients vary depending on the crack width, crack density, and concrete surface density as described above, the wavelet coefficients related to the crack density and concrete surface density are calculated for each gradation using pseudo-created data, and a wavelet coefficient table is created. For example, the wavelet coefficients (thresholds) corresponding to the two contrasting densities (one density can be assumed to be the concrete surface density and the other density to be the crack density) are uniquely determined by referring to the wavelet coefficient table.

The binary image creation unit 510 executes a binary process that compares each pixel of the wavelet image created by the wavelet image creation unit 506 with a wavelet coefficient that serves as a threshold value created by the wavelet coefficient table creation unit 508. For example, if the wavelet coefficient value of each pixel that constitutes the wavelet image is greater than the threshold value of the wavelet coefficient table, the pixel is determined to be a crack and assigned a value of 1, and if the wavelet coefficient value is less than the threshold value, the pixel is determined to not be a crack and assigned a value of 0. Therefore, the binary image is an image in which each pixel is represented by either 0 or 1.

The crack image creation unit 512 creates a crack image by performing contour tracing processing on the binary image and then performing thinning processing.

The contour tracing process starts from an arbitrary pixel (a pixel that is determined to be a crack), and if an adjacent pixel is a crack, it connects to the starting pixel, and if another adjacent pixel is a crack, it connects both of them again, and ends when it finally closes to the starting pixel (for example, when the first pixel, the second pixel, ..., the n-1th pixel, the nth pixel, and the first pixel are connected in that order) or when there is no further crack to connect to. This contour tracing process creates appropriate crack lines, such as crack lines that close in a loop or crack lines that extend linearly with multiple bends. At this time, by setting a predetermined value for the minimum number of pixels that can be connected, all pixels below this set number are deemed not to be cracks and can be removed from the crack display on the screen.

This contour tracing process labels blocky shapes with geometric connectivity into individual cracks, and large noise is removed during this labeling process.

Following the contour tracing process is a thinning process, which reduces the crack to a single crack with a width of, for example, one pixel, consisting of its center line. This process allows for the removal of small noise.

In this way, by performing contour tracing on the binarized image and then thinning, a crack image with multiple labeled cracks is created.

The crack image created is a highly accurate crack image on a pixel-by-pixel basis. However, even if the end points of cracks are physically separated, cracks that can be determined to be essentially continuous are preferably output as cracks with the same label, as this allows for crack detection and output in the same way as when an inspector actually inspects and determines the cracks. Therefore, the segmentation image creation unit 514 creates a segmentation image by connecting the end points of cracks that are determined to be essentially continuous among the multiple separated cracks in the crack image.

More specifically, the segmentation image creation unit 514 has an end point distance determination unit 520, a determination unit 522, and a drawing unit 524.

The end point distance determination unit 520 determines the distance between mutually adjacent end points of multiple separated cracks in a crack image. For example, even if an end point of one crack intersects with the middle of another adjacent crack, the end points of both are separated, so the end point distance is determined even in such a separated state.

The determination unit 522 compares the distance between the end points with a threshold value for the distance between the end points, and if the distance between the end points is equal to or less than the threshold value, it determines that the two cracks whose distance between the end points is identified are essentially continuous cracks and that both are cracks with the same label.

Here, the method for setting the threshold for the distance between the endpoints is to identify a distance (e.g., 3 cm) at which there is a high probability (e.g., 70% or higher probability) that the cracks are of the same label even if they are separated from each other, based on accumulated past crack detection results on the concrete surface, and set the threshold. In addition, the width at the endpoints of the cracks that are separated from each other may also be taken into consideration, and the threshold may be set based on both the width of the endpoints and the distance.

Furthermore, the threshold value can also be set by the machine learning unit 530. The machine learning unit 530 may set the threshold value based on a machine-learned function, etc., using a large amount of past data on the combination of cracks (including the linearity of both cracks and the crack width at the end points) when the cracks that are separated from each other can be determined to be essentially continuous cracks, and the distance between the end points at that time.

The machine-learned function can be created using existing technology, such as supervised learning using a neural network, but the creation method is not particularly limited. By repeating machine learning in the machine learning unit 530, an optimal threshold value corresponding to the spaced crack pattern is learned and set based on various crack patterns in various concrete surface images.

As described above, the determination unit 522 creates a segmentation image by comparing the distance between the endpoints of the separated cracks with a threshold value.

The segmentation image is created here by automatically connecting the cracks in the drawing unit 524. It may also be created by an analyst using image editing software (handwriting software) to connect the cracks. In the latter case, as a result of the judgment by the judgment unit 522, the endpoints of the cracks that should be connected to each other are displayed on the screen, allowing the analyst to connect the displayed endpoints.

As described above, in addition to the judgment method in which the judgment unit 522 only compares the distance between the endpoints with a threshold value, when it is judged that the distance between the endpoints is equal to or less than the threshold value by comparing the distance between the endpoints with the threshold value, it can also execute a judgment method in which it subsequently verifies the directionality (vector) near the endpoints of both cracks, and if it is judged that both vectors have continuity, it judges that they are cracks with the same label. That is, the judgment unit 522 can make a judgment using either of these methods, and the analyst can selectively select the judgment algorithm in the judgment unit 522. For example, a simple judgment can be made by a judgment based only on a comparison of the distance between the endpoints with a threshold value, and a detailed judgment can be made by a judgment that identifies continuity based on vectors near the endpoints of both cracks in addition to comparing the distance between the endpoints with the threshold value.

Even if the distance between the endpoints is below the threshold, if it is not possible to determine that the cracks are clearly of the same label based on the vectors near the endpoints of both cracks, then it is common for an inspector not to determine that the two cracks are of the same label when actually inspecting them. Therefore, this detailed judgment makes it possible to create a segmentation image with higher accuracy. The method of determining continuity based on vectors near the endpoints of both cracks in the detailed judgment is explained in detail below.

The output unit 516 selectively outputs either the crack image or the segmentation image, or both. The analyst can select the output method as desired.

In addition, in the crack distribution map, which is an example of the output, the crack width (maximum crack width) of each crack is added to the side of the crack, making it possible to check the distribution of each crack on the concrete surface and the damage condition of each crack on a single screen.

As for identifying the maximum crack width in a segmentation image, as explained in the section on verification experiments below, it has been demonstrated that by deleting the top 10% of cracks with the largest crack widths in the segmentation image and determining the maximum crack width, it is possible to identify a maximum crack width that is close to the actual measurement value by an inspector.

The output unit 516 outputs crack distribution diagrams such as crack images and segmentation images, as well as crack information regarding the crack length for each crack width, average crack width, crack density, total crack length, etc.

<Crack detection method>
Next, an example of a crack detection method according to an embodiment will be described with reference to Fig. 4 to Fig. 13. Here, Fig. 4 is a flowchart showing an example of a crack detection method according to an embodiment, and shows a process flow in the crack detection device 300.

In step S700, an input image is created in the input image creation unit 504 based on the image data input into the input unit 502 (input image creation process).

For example, if the input image requires brightness correction processing, the brightness correction processing is performed to create the input image, and if brightness correction processing is not required, the captured image is used as is as the input image. Of the 256 brightness levels, a brightness that the analyst determines to be optimal is set within a range of approximately 120 to 160, including the median value of 128. If the brightness of the captured image does not match the set brightness, a brightness correction processing is performed on the captured image, and an input image with the set brightness is created.

In step S702, an input image is input to the wavelet image creation unit 506, and a wavelet transform process is performed on the input image to create a wavelet image. At this time, the wavelet image may be created by extracting assumed cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on the concrete surface, drawing along the assumed cracks to create assumed crack drawing lines, and performing wavelet transform process on the drawing-containing image including the assumed crack drawing lines. Alternatively, a path image including paths added along assumed crack lines that are assumed to be cracks is created for the input image, and wavelet transform process is performed on the path image to create the wavelet image (wavelet image creation process).

Now, we will explain the wavelet transform process. Figure 5A is a diagram showing an example of the relationship between an input image and a local region in wavelet transform process. Also, Figure 5B is a diagram showing an example of the relationship between a local region and a pixel of interest in wavelet transform process. Wavelet transform is performed in local region 3, which is the center of wide area 2 in input image 1, and cracks are detected in the center of local region 3. Cracks in input image 1 are detected by translating wide area 2 up, down, left and right all over input image 1.

Figure 5B is an enlarged view of local region 3. In the illustrated embodiment, a crack is determined at the center of, for example, 3 x 3 nine pixels (eight neighboring pixels 31, 31, ... and a pixel of interest 32 located in the center). Note that the wavelet coefficient is calculated for local region 3 in Figure 5A. The formula for calculating the wavelet coefficient by performing a wavelet transform using a wavelet function (mother wavelet function) is shown below.

Here, f(x, y) indicates the input image (where x and y are arbitrary coordinates in the two-dimensional input image), Ψ indicates the mother wavelet function (Gabor function), and (x ₀ , y ₀ ) indicates the translation amount of Ψ. Furthermore, a ^k indicates the expansion or contraction of Ψ (where a ^k is the inverse of the frequency and is a value indicated by an integer k for the frequency width for calculations for several frequency regions), f ₀ indicates the center frequency, and σ indicates the standard deviation of the Gaussian function. Furthermore, θ indicates the rotation angle representing the traveling direction of the wave, and (x', y') indicates the coordinates obtained by rotating (x, y) by the angle θ.

The cumulative total C(x ₀ , y ₀ ) of the wavelet coefficient Ψ for a plurality of θ and k calculated using Equation 1 is given by Equation 4 below.

The above parameters can be set arbitrarily, for example, σ can be set to 0.5 to 2, a ^k can be set to 0 to 5, f ₀ can be set to 0.1, and the rotation angle can be set to 0 to 180 degrees. The translation amount (x ₀ , y ₀ ) in Equation 4 corresponds to the position of the pixel of interest, and the continuous amount of wavelet coefficients (C(x ₀ , y ₀ )) can be calculated by sequentially moving the position of the pixel of interest.

After calculating the wavelet coefficients for all pixels that make up local region 3 based on the above formula, the wavelet coefficients are similarly calculated for all pixels in wide region 2 that is created by shifting the pixel of interest one position to the left or right or up or down.

When creating a wavelet image, the wavelet coefficient table is created by the wavelet coefficient table creation unit 508. In creating the wavelet coefficient table, wavelet coefficients are calculated for a pseudo image consisting of two contrasting densities that have no relationship to the input image. For example, as shown in FIG. 6, wavelet coefficients are obtained for a pseudo image consisting of a background color a assumed to be a concrete surface (for example, the R, G, and B background colors are 255, 255, and 255) and line segments b1 to b5 assumed to be cracks. Here, the line widths of the line segments b1 to b5 vary in order from 1 pixel to 5 pixels, and each line segment has three types of densities (for example, the line segment b1 varies in order of density from b11 (black), b12 (light black), and b13 (gray)). A bird's-eye view of the wavelet coefficients calculated by performing a wavelet transform on this pseudo image is shown in FIG. 7. In FIG. 7, the X-axis represents the width of the line segment, the Y-axis represents the density of the color of the line segment, and the Z-axis represents the wavelet coefficient. At the same time, a wavelet coefficient table as shown in FIG. 8 is created by combining two contrasting densities at 256 gradations from 0 to 255. The wavelet coefficient table may be created at any time up to the creation of a binary image, which will be described later. For example, it may be created after the captured image is input, or it may be created in parallel with the creation of a wavelet image, or it may be created after the wavelet image is created and before the creation of a binary image.

In step S704, the binary image creation unit 510 creates a binary image (binary image creation process).

In the wavelet coefficient table, the wavelet coefficient corresponding to the average density of the neighboring pixels in the local region and the density of the pixel of interest is set as the threshold value for the wavelet coefficient. If the wavelet coefficient of the pixel of interest is larger than the threshold, the pixel of interest is determined to be a crack (e.g., white on the screen); if it is smaller, the pixel of interest is determined not to be a crack (e.g., black on the screen). A binary image is created by performing a comparison operation between the wavelet coefficient of the pixel of interest and the threshold while changing the local region and the pixel of interest.

In step S706, the binarized image is input to the crack image creation unit 512, blocky figures with geometric connectivity are labeled into individual cracks, and a crack image with multiple labeled cracks is created (crack image creation process).

As described above, in the crack image creation process, a contour tracing process is performed on the binarized image, followed by a thinning process to create a crack image while removing both large and small noise.

In step S708, the crack image is input to the segmentation image creation unit 514, and a segmentation image is created by connecting the end points of multiple cracks that are separated from each other in the crack image and that are determined to be essentially continuous cracks (segmentation image creation process).

The segmentation image creation process includes an end point distance determination process, a determination process, and a drawing process.

In the end point distance identification process, the end point distance identification unit 520 identifies the end point distance between the end points of mutually separated cracks. Since a crack image contains a large number of mutually separated cracks (combinations), the distance between the end points of all mutually separated cracks is identified.

In the determination process, the determination unit 522 compares the distance between the endpoints with a threshold value for the distance between the endpoints, and if the distance between the endpoints is equal to or less than the threshold value, it determines that the two cracks whose distance between the endpoints is identified are essentially continuous cracks and are cracks with the same label. As described above, there are several methods for setting this threshold, such as a method of setting it based on past performance, or a method of setting it by having the machine learning unit 530 learn past performance by machine learning.

In the drawing process, a segmentation image is created by connecting the end points of two cracks that have been determined to be cracks of the same label. As described above, the drawing method in the drawing process is to automatically connect the cracks in the drawing unit 524. Note that the analyst may also use handwriting software to connect the cracks.

The above-mentioned judgment method in the judgment process is a simple judgment in which only the distance between the endpoints is compared with a threshold value, but when the distance between the endpoints is judged to be equal to or less than the threshold value by comparing the distance between the endpoints with the threshold value, a detailed judgment can be performed in which the directionality (vector) near the endpoints of both cracks is verified, and if it is determined that both vectors have continuity, it is determined that the cracks have the same label. The analyst can selectively select the judgment algorithm in the judgment unit 522.

Here, the contents of the detailed judgment will be described with reference to Figs. 9 to 11. Fig. 9(a) is a schematic diagram showing an example in which cracks separated from each other are judged to be essentially continuous cracks, and Figs. 9(b) and 9(c) are both schematic diagrams showing an example in which cracks separated from each other are judged not to be essentially continuous cracks. Also, Figs. 10(a), (b), and (c) are all schematic diagrams showing an example of judgment that is further performed in the judgment unit using Fig. 9(a) when the distance between the end points is equal to or less than a threshold value. Furthermore, Figs. 11(a) and 11(b) are schematic diagrams showing an example of judgment that is further performed in the judgment unit using Figs. 9(b) and 9(c), respectively, when the distance between the end points is equal to or less than a threshold value.

In Figures 9(a) to 9(c), the distance t1 between the endpoints P1, P2 of the mutually separating cracks C1, C2, the distance t2 between the endpoints P3, P4 of the cracks C3, C4, and the distance t3 between the endpoints P5, P6 of the cracks C5, C6 are all below a set threshold.

In FIG. 9(a), vectors V1 and V2 near the end points of cracks C1 and C2 (for example, within a range of about 5 pixels from the inside of the crack to end points P1 and P2) are parallel or approximately parallel to each other (shifted about 30 to 45 degrees from parallel), and it can be judged from the outside that end points P1 and P2 extend in a direction in which they may intersect with each other. Therefore, as a method for judging whether or not cracks are of the same label, a method is applied in which an endless outline range is set at the end points of one crack, and if the end point of the other crack falls within this set endless outline range, both cracks are judged to be cracks of the same label, and if they do not fall within the endless outline range, they are judged not to be cracks of the same label.

Examples of endless contour shapes are shown in Figures 10(a) to 10(c). The example shown in Figure 10(a) is an example in which an equilateral triangle (an example of a regular polygon) with one side of a predetermined length r1 is used as the endless contour range, Figure 10(b) is an example in which a square (an example of a regular polygon) with one side of a predetermined length r1 is used as the endless contour range, and Figure 10(c) is an example in which a circle with a diameter of a predetermined length r1 is used as the endless contour range.

The specified length r1 that defines the endless contour range can be set based on the performance of past crack detection methods, from the distance between the endpoints of cracks that can be determined to be cracks of the same label even if they are separated from each other, and can also be set to a value equal to a threshold value, for example.

In Figure 10(a), the vertex of an equilateral triangle is positioned at the endpoint P1 of one crack C1, and an equilateral triangle is set so that a virtual extension line extending vector V1 near the endpoint P1 of one crack C1 intersects with the midpoint of the base of the equilateral triangle, and the contour range A1 of this equilateral triangle is a range of 30 degrees left and right from the extension line of vector V1. In Figure 10(a), because the endpoint P2 of the other crack C2 falls within the contour range A1 of the equilateral triangle, cracks C1 and C2 can be determined to be cracks with the same label.

In contrast, in Figure 10(b), one corner of the square is positioned at the endpoint P1 of one crack C1, and another corner of the square that is diagonally opposite the one corner is positioned on an imaginary extension line of vector V1 near the endpoint P1 of one crack C1 to set a square, and the contour range A2 of this equilateral triangle is a 45-degree range to the left and right of the extension line of vector V1. In Figure 10(b), because the endpoint P2 of the other crack C2 falls within the contour range A2 of the square, cracks C1 and C2 can be determined to be cracks with the same label.

In addition, in Figure 10(c), one point on the circumference of the circle is positioned at the end point P1 of one crack C1, and another point on the circle that is at both ends of the diameter of the circle is positioned on an imaginary extension line extending from the end point P1 of one crack C1 along a vector V1 near the end point, thereby setting the circle, and the circle of the contour range A3 can be set in the direction of the extension line of the vector V1.In Figure 10(c), because the end point P2 of the other crack C2 falls within the contour range A3 of the circle, cracks C1 and C2 can be determined to be cracks with the same label.

The endless outline range may be a polygon other than a square or equilateral triangle, and may be a circle or an ellipse, etc. However, the shape shown in the example is preferred because squares and equilateral triangles are the polygons that are easiest to clearly define the outline range, and circles can also be arranged in the vector direction as in the example shown, making it easy to clearly define the outline range by setting the radius and diameter.

In addition, if an endless contour range is set at the end point of one crack and the end point of the other crack does not fall within it, the opposite case is performed, an endless contour range is set at the end point of the other crack, and it is checked whether the end point of one crack falls within it. If it does, both cracks are determined to be cracks of the same label, and if it does not (in both cases, the end points of the adjacent crack do not fall within the endless contour range), both cracks are determined not to be cracks of the same label.

On the other hand, FIG. 11(a) shows the case where an equilateral triangular contour range is defined for the model in FIG. 9(b), and FIG. 11(b) shows the case where a circular contour range is defined for the model in FIG. 9(c).

In Figures 11(a) and 11(b), the equilateral triangle and circular contour range A1, A3 set for the endpoints P3, P5 of one crack, C3, C5, do not include the endpoints P4, P6 of the other crack, C4, C6, and vice versa, the equilateral triangle and circular contour range A1, A3 set for the endpoints P4, P6 of the other crack, C4, C6, do not include the endpoints P3, P5 of one crack, C3, C5, so it is determined that cracks C3, C4 and cracks C5, C6 are not cracks with the same label.

In this way, by performing a detailed judgment in the judgment process to determine whether or not cracks have the same label, it becomes possible to create a segmentation image with even higher accuracy.

Returning to FIG. 4, in step S710, various images are output by the output unit 516 (output process). Here, FIG. 12 is a photograph showing a pixel-by-pixel crack image output by the output unit 516, and FIG. 13 is a photograph showing a segmentation image output by the output unit 516.

For the crack image in Figure 12, after going through the segmentation image creation process, some of the cracks that are separated from each other are determined to be cracks of the same label and segmented, so that two cracks of the same label with crack widths (maximum crack widths) of 0.1 mm and 0.2 mm are formed in the segmentation image, as shown in Figure 13.

As shown in Figure 13, in the segmentation image, by adding the crack width (maximum crack width) near all cracks, including the segmented cracks, it is possible to check the distribution of each crack on the concrete surface and the damage condition of each crack on a single screen.

In the output process, the analyst can choose to output either the crack image or the segmentation image, or both, to the computer screen.

The crack detection method shown in the figure not only detects and outputs cracks with high accuracy on a pixel-by-pixel basis, but also detects and outputs cracks that are physically separated from each other, but that are essentially related and can be determined to be continuous, as cracks with the same label. The crack detection method shown in FIG. 4 is also a series of processing flows executed in the crack detection device 300, but the crack detection device 300 may be formed by installing a program including this series of processing flows on a computer.

[Verification experiments and their results]
The inventors conducted a verification experiment to verify the accuracy of the crack widths identified by the segmentation images. In this verification experiment, the cracks numbered 1 to 3 on the concrete surface (test specimen 1) shown in Fig. 14 and the concrete surface (test specimen 2) shown in Fig. 15 were segmented, and the maximum crack widths of the cracks corresponding to the numbers 1 to 3 were identified as follows: the maximum crack width after deleting the top 20% of the cracks with the largest crack widths, the maximum crack width after deleting the top 10% of the cracks with the largest crack widths, and the maximum crack width without deletion; and the crack widths of each crack were measured by an observer for comparison with the above.

The results of the verification experiment are shown in Table 1, and Figures 16 and 17 show examples of output in which the maximum crack width is noted after deleting the top 10% of cracks with the largest crack width in the segmentation images of test specimens 1 and 2. Note that in Figures 16 and 17, the crack number and crack width (maximum crack width) are noted as a set next to the crack (for example, the notation 52:0.44 indicates that the crack width of crack number 52 is 0.44 mm).

Table 1 demonstrates that, compared to the actual crack width measured by an observer, the crack width without any deletion has the largest error and is larger than the actual crack width, and that the crack width closest to the actual measurement is obtained when the top 10% is deleted.

Based on the results of this verification experiment, it has been demonstrated that it is preferable to use the maximum crack width after deleting the top 10% as the crack width for each crack to be added to the segmentation image.

The above describes the embodiment of the present invention in detail using the drawings, but the specific configuration is not limited to this embodiment, and even if there are design changes, etc., within the scope that does not deviate from the gist of the present invention, they are included in the present invention.

100: Imaging device 200: Network 300: Crack detection device 301: Data collection unit 302: Data processing unit 303: Data storage unit 504: Input image creation unit 506: Wavelet image creation unit 508: Wavelet coefficient table creation unit 510: Binary image creation unit 512: Crack image creation unit 514: Segmentation image creation unit 516: Output unit 520: End point distance determination unit 522: Judgment unit 530: Machine learning unit 532: Drawing unit 1000: Crack detection system C1 to C6: Crack P1 to P6: End points t1 to t3: End point distance V1 to V6: Vector near end points A1 to A3: Contour range (endless contour range)
r1: Predetermined length

Claims (6)

an input image creation step of inputting a photographed image of a concrete surface containing cracks into a computer to create an input image;
a wavelet image creation process for performing a wavelet transform process on the input image, or for extracting assumed cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on a concrete surface, drawing along the assumed cracks to create assumed crack drawing lines, and for performing a wavelet transform process on a drawing-containing image including the assumed crack drawing lines, or for performing a wavelet transform process on a path image including paths added along assumed crack lines that are assumed to be cracks to the input image, to create a wavelet image in which each pixel has a wavelet coefficient;
a binary image creation step of binarizing each pixel of the wavelet image by using a wavelet coefficient table that is a threshold value of the wavelet coefficient to create a binary image;
A crack image creation process for labeling each crack with a block-like figure having geometric connectivity in the binarized image and creating a crack image including a plurality of labeled cracks;
A segmentation image creation process is provided for creating a segmentation image by connecting end points of cracks that are determined to be essentially continuous cracks among the multiple separate cracks in the crack image,
The segmentation image creation step includes:
An end-to-end distance determination step of determining an end-to-end distance between end points of the cracks that are separated from each other;
a determination step of comparing a threshold value for the distance between the end points with the distance between the end points, and determining that the two cracks whose distance between the end points is specified are essentially continuous cracks and are cracks with the same label when the distance between the end points is equal to or less than the threshold;
A drawing process for creating the segmentation image by connecting the end points of two cracks that have been determined to be cracks of the same label ,
In the determination step,
A crack detection method characterized in that, when the distance between the endpoints is equal to or less than the threshold value, an endless contour range is set starting from an endpoint of one of the cracks, and when the endpoint of the other crack falls within the endless contour range, the two cracks whose endpoint distance has been identified are determined to be cracks of the same label. The crack detection method described in claim 1, characterized in that the endless contour range is a regular polygon with a side of a predetermined length or a circle with a diameter of a predetermined length set in the extension direction of the crack, starting from an end point of one of the cracks, and is the contour range of the regular polygon or circle. An output step is further included.
A crack detection method according to claim 1 or 2 , characterized in that when the segmentation image is created, either or both of the crack image and the segmentation image are selectively output. an input image creation unit that creates an input image based on a photographed image of a concrete surface containing cracks that is input to a computer;
a wavelet image creation unit that performs a wavelet transform process on the input image, or extracts assumed cracks that are assumed to be cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on a concrete surface, draws along the assumed cracks to create assumed crack drawing lines, and performs a wavelet transform process on a drawing-containing image that includes the assumed crack drawing lines, or performs a wavelet transform process on a path image that includes paths added along assumed crack lines that are assumed to be cracks on the input image, and creates a wavelet image in which each pixel has a wavelet coefficient;
a binary image creation unit that uses a wavelet coefficient table that is a threshold value of wavelet coefficients to binarize each pixel of the wavelet image to create a binary image;
A crack image creation unit that labels each crack with a block-like figure having geometric connectivity in the binarized image and creates a crack image including the labeled cracks;
A segmentation image creation unit that creates a segmentation image by connecting end points of cracks that are determined to be essentially continuous cracks among the multiple separate cracks in the crack image,
The segmentation image creation unit includes:
An end point distance specification unit that specifies an end point distance between end points of the cracks that are separated from each other;
a determination unit that compares a threshold value for the distance between the end points with the distance between the end points, and when the distance between the end points is equal to or less than the threshold value, determines that the two cracks whose distance between the end points is specified are essentially continuous cracks and are cracks with the same label ;
In the determination unit,
A crack detection device characterized in that, when the distance between the endpoints is less than or equal to the threshold value, an endless contour range is set starting from an endpoint of one of the cracks, and when the endpoint of the other crack falls within the endless contour range, the two cracks whose endpoint distance has been identified are determined to be cracks of the same label. Further comprising an output section,
The crack detection device according to claim 4 , characterized in that when the segmentation image is created, either or both of the crack image and the segmentation image are selectively output. A program for causing a computer to execute the following processes for detecting cracks on a concrete surface,
an input image creation step of inputting a photographed image of a concrete surface containing cracks into a computer to create an input image;
a wavelet image creation process for performing a wavelet transform process on the input image, or for extracting assumed cracks from the input image based on learning information obtained by having a computer learn the characteristics of cracks on a concrete surface, drawing along the assumed cracks to create assumed crack drawing lines, and for performing a wavelet transform process on a drawing-containing image including the assumed crack drawing lines, or for performing a wavelet transform process on a path image including paths added along assumed crack lines that are assumed to be cracks to the input image, to create a wavelet image in which each pixel has a wavelet coefficient;
a binary image creation step of binarizing each pixel of the wavelet image by using a wavelet coefficient table that is a threshold value of the wavelet coefficient to create a binary image;
A crack image creation process for labeling each crack with a block-like figure having geometric connectivity in the binarized image and creating a crack image including a plurality of labeled cracks;
A segmentation image creation process is provided for creating a segmentation image by connecting end points of cracks that are determined to be essentially continuous cracks among the multiple separate cracks in the crack image,
The segmentation image creation step includes:
An end-to-end distance determination step of determining an end-to-end distance between end points of the cracks that are separated from each other;
a determination step of comparing a threshold value for the distance between the end points with the distance between the end points, and determining that the two cracks whose distance between the end points is specified are essentially continuous cracks and are cracks with the same label when the distance between the end points is equal to or less than the threshold value ;
In the determination step,
A program characterized in that, when the distance between the endpoints is equal to or less than the threshold value, the program further sets an endless contour range starting from an endpoint of one of the cracks, and when the endpoint of the other crack falls within the endless contour range, the program determines that the two cracks whose endpoint distance has been identified are cracks of the same label.

Images