JP6629685B2 - Crack analysis method and crack analysis system - Google Patents

Description

  The present invention relates to a crack analysis method and a crack analysis system for a crack generated in an inspection target such as rock or concrete.

  2. Description of the Related Art As a method of analyzing a crack generated in an inspection target such as a bedrock or concrete, various methods are conventionally known. As a conventional method, for example, at a construction site of a tunnel, there is a method in which a crack generated in a face of the tunnel is described by a sketch by a site worker. The site worker sketches the face while looking at the face as a whole, thereby grasping the state of the cracks generated in the face as a whole.

  Further, Japanese Patent Application Laid-Open No. 10-331575 describes a geological prediction method in front of a tunnel face used when observing a tunnel face at a tunnel excavation site, and an image processing system for performing a geological prediction. . The image processing system includes a digital camera that captures an image of a face, an image memory that stores digital image data acquired by the digital camera, and an image processing operation unit that performs image processing on the image data stored in the image memory. And

  The image processing operation unit extracts a crack portion and a weatherable brittle portion from the image data based on a density distribution. Further, the image processing operation unit obtains a fractal dimension by performing a predetermined operation on the extracted image data. The fractal dimension is used in a grayscale image to quantitatively represent the complexity of a tissue such as undulations and cracks on the surface of a face.

JP-A-10-331575

  In the above description of the crack by the sketch of the site worker, the quality of the sketch varies depending on the skill level of the site worker, so that the accuracy of grasping the state of the crack varies. On the other hand, in the image processing system of the above-mentioned publication, an index relating to the complexity of unevenness on the surface of the face is output as a fractal dimension. Therefore, it is possible to output an index relating to the complexity of the face of the face as a numerical value.

  However, as described above, the field worker tries to grasp the state of the crack as a whole, and therefore, there is a situation in which the output by the numerical value of the fractal dimension is not suitable for the field worker. That is, the field worker cannot easily grasp the state of the occurrence of cracks by looking at the output result of the fractal dimension. Therefore, it is required to output data that allows a site worker to easily grasp the state of the crack.

  An object of the present invention is to provide a crack analysis method and a crack analysis system that allow a site worker to easily grasp the state of occurrence of a crack.

The crack analysis method according to the present invention is a crack analysis method for analyzing a crack generated in an inspection target, acquires an image of the inspection target, and converts the image into a crack pixel indicating a portion where a crack is generated. A first step of binarizing a non-cracked pixel indicating a portion where no crack has occurred, and, for each rotation angle with respect to the reference direction, while rotating the image around an axis orthogonal to the image with respect to the reference direction, A second step of calculating a degree in which the extension direction of the crack pixel coincides with the reference direction, and determining a direction in which the calculated degree is rotated by the highest rotation angle in the opposite direction from the reference direction as a direction in which the crack extends. A third step of displaying, and a fourth step of displaying a direction in which the determined crack extends.
The second step is a step of dividing the image into a plurality of divided images, and calculating a ratio of the number of the divided images in which the crack pixels extend from one end to the other end in the reference direction with respect to the total number of the divided images. And

  In the crack analysis method according to the present invention, the image is rotated and the degree of coincidence of the extension direction of the crack pixel with respect to the reference direction is calculated for each rotation angle with respect to the reference direction. Then, a rotation angle having a high degree of coincidence is specified, a direction rotated in the opposite direction from the reference direction by the specified rotation angle is determined as a direction in which the crack extends, and the direction is displayed. Therefore, by performing the image processing for rotating the acquired image, the direction in which the crack extends in the inspection object can be visually grasped. Therefore, it is possible to output, to a field worker who manages the inspection object, data that can easily grasp the state of occurrence of cracks. As described above, the direction in which the crack extends in the inspection object is displayed, so that the field worker can intuitively grasp the state of the crack occurrence.

Further, the second step is a step of dividing the image into a plurality of divided images, and a step of calculating a ratio of the number of the divided images in which the crack pixels extend from one end to the other end in the reference direction with respect to the total number of the divided images. If, you have a. Therefore , the degree of coincidence of the extension direction of the crack pixel with respect to the reference direction can be calculated as a numerical value. Since the calculated ratio is a ratio of the number of divided images in which the crack pixels extend from one end to the other end in the reference direction, it is a simple algorithm compared with a predetermined calculation such as a fractal dimension. Therefore, the calculation process in each step can be speeded up and the data output can be speeded up, so that the data indicating the state of occurrence of cracks can be quickly output to the site worker.

  Further, the step of dividing and the step of calculating the ratio are performed a plurality of times while changing the number of divided images, and a step of calculating a relational expression between the size and the ratio of the divided image, and a step of calculating an integral value of the relational expression. And calculating. In this case, the direction in which the crack extends can be determined with higher accuracy by calculating the degree to which the extension direction of the crack pixel matches the reference direction as an integral value. Therefore, the extending direction of the crack to be displayed can be output as data that matches the actual feeling of the site worker.

  Further, after the third step, in a state where the image is rotated by the rotation angle with respect to the reference direction, the total number of pixels located on the imaginary line extending in the direction orthogonal to the reference direction is determined by the number of crack pixels located on the imaginary line. The method may further include a step of calculating an interval between the cracks by dividing the distance. As described above, by calculating the interval between the cracks, the density of the cracks generated in the inspection object can be grasped. Therefore, the state of occurrence of cracks can be grasped with higher accuracy.

  The second, third, and fourth steps are performed for each of a plurality of regions of the inspection object. In the fourth step, for the region in which the highest degree in the third step exceeds the first threshold, Only the direction in which the crack extends may be displayed. In this case, the direction in which the cracks extend is displayed only in the region where the crack pixels extend in a certain direction. Therefore, by selectively extracting cracks that need to be grasped, the field worker can grasp the cracks more reliably.

  Further, the second step, the third step, the fourth step, and the step of calculating the interval are performed for each of a plurality of areas of the inspection object, and in the fourth step, the interval calculated in the step of calculating the interval is the fourth interval. The direction in which the crack extends may be displayed only for an area that is equal to or less than two thresholds. In this case, the direction in which the cracks extend is displayed only in the region where the cracks are formed with a smaller gap and higher density. Therefore, by selectively extracting cracks that need to be grasped, the field worker can grasp the cracks more reliably.

The crack analysis system according to the present invention is a crack analysis system that analyzes a crack generated in an inspection object, acquires an image of the inspection object, and generates an image with a crack pixel indicating a portion where a crack is generated. An image binarization unit that binarizes a non-crack pixel indicating a portion where no crack has occurred, and a rotation angle with respect to the reference direction while rotating the image around an axis orthogonal to the image with respect to the reference direction. In addition, a degree calculating unit that calculates the degree at which the extension direction of the crack pixel matches the reference direction, and the direction in which the calculated degree is rotated from the reference direction by the highest rotation angle in the opposite direction to the direction in which the crack extends And a display unit that displays the direction in which the determined crack extends . The degree calculation unit is a connection ratio calculation unit that divides the image into a plurality of divided images. Part is based on the total number of divided images. Crack pixel from the reference direction of the one end to the other end to calculate the ratio of the number of extending divided images.

  In the crack analysis system according to the present invention, the degree calculation unit calculates the degree of coincidence of the extension direction of the crack pixel with respect to the reference direction while rotating the image. The determining unit specifies a rotation angle having a high degree of coincidence, and determines a direction in which the specified rotation angle is rotated in the opposite direction from the reference direction as a direction in which the crack extends. Then, the display unit displays the direction in which the determined crack extends. Therefore, the degree calculation unit rotates the image, and the determination unit determines the direction in which the crack extends, so that the direction in which the crack has occurred in the inspection object can be grasped. Therefore, it is possible to output, to a field worker who manages the inspection object, data that can easily grasp the state of occurrence of cracks. As described above, the direction in which the crack extends in the inspection object is displayed, so that the field worker can intuitively grasp the state of the crack occurrence.

  ADVANTAGE OF THE INVENTION According to this invention, a field worker can grasp | ascertain the generation | occurrence | production state of a crack easily.

It is a block diagram showing a crack analysis system concerning an embodiment of the present invention. It is a figure which shows the image data of the inspection object (face of a tunnel) image | photographed and taken in by the camera. FIG. 3 is a diagram illustrating image data obtained by cutting out a part of the image data in FIG. 2. FIG. 4 is a diagram illustrating an image obtained by binarizing the image data of FIG. 3 into crack pixels and non-crack pixels. FIG. 5A is a diagram illustrating a divided image obtained by dividing the image of FIG. 4 into four parts. (B) is a graph showing a relationship between the size of the divided image and the connection ratio. (A) is a figure which shows the division | segmentation image obtained by dividing the image of FIG. 4 into nine. (B) is a graph showing a relationship between the size of the divided image and the connection ratio. (A) is a figure which shows the division | segmentation image obtained by dividing the image of FIG. 4 into 16 parts. (B) is a graph showing a relationship between the size of the divided image and the connection ratio. 9 is a graph showing a relationship between a size of a divided image and a connection ratio. (A) is a figure which shows a crack pixel and a non-crack pixel when it rotates 50 degrees clockwise from a reference direction. (B) is a figure which shows the crack pixel and the non-crack pixel which are not rotated with respect to a reference direction. (C) is a diagram showing cracked pixels and non-cracked pixels when rotated clockwise by 20 ° from the reference direction. It is a figure showing the step of calculating a crack interval. It is a figure showing a plurality of fields of an inspection subject (face of a tunnel). It is a figure which shows the direction where the determined crack extends for every area | region. It is a figure which shows the step which displays the direction in which a crack extends from each direction of FIG. It is a figure which shows the step which displays the direction in which a crack extends following FIG. It is a flowchart which shows each step of the crack analysis method which concerns on embodiment of this invention.

  Hereinafter, an embodiment of a crack analysis method and a crack analysis system according to the present invention will be described with reference to the drawings. In the following description, the same or corresponding elements have the same reference characters allotted, and overlapping description will be omitted.

  First, a crack analysis system according to the present embodiment will be described. FIG. 1 is a block diagram illustrating a crack analysis system 1 according to the present embodiment, and FIG. 2 is a diagram illustrating image data of a tunnel face T to be analyzed by the crack analysis system 1. The tunnel face T is, for example, a cracked rock, and the stability of the tunnel constructed at the tunnel face T is easily affected by the state of the crack. Therefore, it is important to perform appropriate evaluation of the bedrock of the tunnel face T by grasping and managing the state of the cracks in the tunnel face T when performing construction of the tunnel. In the present embodiment, a description will be given of a crack analysis system 1 that uses a tunnel face T as an inspection target and analyzes a crack generated in the tunnel face T.

  The crack analysis system 1 provides, for example, technical support to a site clerk who sketches a crack in the tunnel face T. In other words, the crack analysis system 1 has a function of displaying the predominant direction of the crack (the direction in which the crack extends) of the tunnel face T. It is possible to make a sketch based on. Further, the predominant direction of the crack displayed by the crack analysis system 1 is easy to understand for the site worker, and can be replaced with the above sketch.

  As shown in FIG. 1, a crack analysis system 1 includes an image acquisition unit 2 that acquires a captured image of a tunnel face T, an image processing unit 10 that performs image processing of a captured image acquired by the image acquisition unit 2, A monitor 3 (display unit) for displaying an image relating to the predominant direction of the crack generated by the image processing unit 10.

  The image acquisition unit 2 is, for example, a digital camera, and acquires digital data of an image of the tunnel face T. Various types can be used as the image acquisition unit 2. The image acquisition unit 2 may include, for example, a camera for taking a silver halide photograph and a scanner, and various cameras capable of acquiring an image of the tunnel face T can be used.

  The monitor 3 displays an image of the crack generated by the image processing performed by the image processing unit 10. The monitor 3 is, for example, a display of a personal computer. Note that as the monitor 3, various display devices can be used as long as the display can display an image of the crack.

  The image processing unit 10 is capable of communicating with the image acquisition unit 2 and the monitor 3. The image processing unit 10 receives the digital data of the image of the tunnel face T acquired by the image acquisition unit 2. The image processing unit 10 includes an image binarization unit 11, an image extraction unit 12, a connection ratio calculation unit 13, an image rotation unit 14, an angle detection unit 15, an interval calculation unit 16, and a display control unit 17. The image processing unit 10 is, for example, a general-purpose personal computer, and includes a CPU (Central Processing Unit), a ROM (Read Only Memory), and a RAM (Random Access Memory). Each function of the image processing unit 10 is realized by loading a program stored in the ROM into the RAM and executing the program by the CPU.

  The image binarization unit 11 receives the digital data of the image of the tunnel face T, and determines the lower left part A1, the lower right part A2, and the top part A3 with respect to the received digital data of the image as shown in FIG. To create an arc-shaped analysis region D. The outer edge of the analysis area D corresponds to the outer edge of the tunnel face T.

  The image binarization unit 11 performs black and white binarization on an image inside the analysis area D. In this embodiment, a white pixel in FIG. 2 is a crack pixel P1 indicating a portion of the tunnel face T where a crack has occurred, and a black pixel in FIG. 2 is a non-crack pixel indicating a portion where the tunnel face T has no crack. P2.

  The image extracting unit 12 extracts an arbitrary part from the analysis area D generated by the image binarizing unit 11, as shown in FIG. That is, the image extracting unit 12 extracts the region R of the tunnel face T from the image binarized by the image binarizing unit 11. Although the region R shown in FIG. 3 has a circular shape, the shape and size of the region R extracted by the image extracting unit 12 are not limited to the example shown in FIG.

  Hereinafter, an example in which the region R is a square as shown in FIG. 4 will be described. The connection ratio calculation unit 13 (degree calculation unit) calculates the connection ratio in order to determine the dominant direction of the crack in the tunnel face T. The connection rate indicates the continuity of the crack in the region R extracted by the image extraction unit 12 in the vertical direction (reference direction). For example, assuming that the image G of the region R is composed of 20 × 20 pixels, in the image G, the crack pixel P1 does not extend from one end (upper end) to the other end (lower end) in the vertical direction. , The connection rate of the image G is 0. In FIG. 4, the crack pixel P1 and the non-crack pixel P2 have a square shape, but may have a shape other than a square, such as a rectangle.

  The connection ratio calculation unit 13 divides the image G into a plurality of divided images B, as shown in FIG. FIG. 5A shows an example in which a region R is divided into four (2 × 2) divided images B. The connection ratio calculation unit 13 calculates a ratio of the number of the divided images B in which the crack pixels P1 extend from the upper end to the lower end to the total number of the divided images B. This ratio corresponds to the connection ratio.

  In the example shown in FIG. 5A, the divided image B in which the crack pixel P1 extends from the upper end to the lower end is only one upper left divided image B. Further, the connection ratio calculation unit 13 divides the image G into four (2 × 2) divided images B. The connection ratio calculation unit 13 calculates the connection ratio as 1/4, with the total number of the divided images B being 4 and the number of the divided images B in which the crack pixels P1 extend from the upper end to the lower end being 1. The connection ratio calculation unit 13 stores the relationship between the size of the divided image B and the connection ratio, as shown in FIG. In the present embodiment, the size of the divided image B indicates the ratio of the number of pixels of one side of the divided image B to the number of pixels (20) of one side of the image G. In the example shown in FIG. 10/20).

  As shown in FIG. 6A, the connection ratio calculation unit 13 divides the image G into nine (3 × 3) divided images B. Then, the connection ratio calculation unit 13 sets the total number of the divided images B to 9, sets the number of the divided images B in which the crack pixels P1 extend from the upper end to the lower end to 3, and calculates the connection ratio as 3/9. Thereafter, the connection ratio calculation unit 13 stores the relationship between the size of the divided image B and the connection ratio, as shown in FIG. At this time, the size of the divided image B is 1/3.

  As shown in FIG. 7A, the connection ratio calculation unit 13 divides the image G into 16 (4 × 4) divided images B. At this time, the connection rate calculation unit 13 calculates the connection rate as 8/16, where the total number of the divided images B is 16, the number of the divided images B in which the crack pixels P1 extend from the upper end to the lower end is 8, and the connection rate is 8/16. Then, the connection ratio calculation unit 13 stores the relationship between the size of the divided image B and the connection ratio, as shown in FIG. 7B.

  As described above, the connection ratio calculation unit 13 performs the division of the image G into the plurality of divided images B and the calculation of the connection ratio a plurality of times while changing the number of the divided images B. Then, as illustrated in FIG. 8, the connection rate calculation unit 13 performs, for example, an approximate expression represented by Y = AX + B (Y: connection rate, X: size of the divided image B, A, B: coefficients). , The relational expression between the size of the divided image B and the connection ratio is calculated. Note that the relational expression calculated by the connection ratio calculation unit 13 is not limited to the linear expression as described above, and can be appropriately changed. For example, the connection ratio calculation unit 13 can use various relational expressions such as a linear approximation curve, a polynomial approximation curve, a logarithmic approximation curve, a power approximation curve, an exponential approximation curve, and a moving average.

  After calculating the relational expression as described above, the connection ratio calculation unit 13 integrates the relational expression and calculates the area S of the graph shown in FIG. 8 as an integral value. This area S (integral value) indicates the degree to which the extending direction of the crack pixel P1 coincides with the vertical direction. That is, the larger the area S, the more cracks extending in the vertical direction, and the smaller the area S, the fewer cracks extending in the vertical direction (there are many cracks extending in other directions, or Itself is small).

  As shown in FIG. 9, the image rotation unit 14 (degree calculation unit) rotates the image G of the region R extracted by the image extraction unit 12. The image rotation unit 14 rotates the image G about an axis perpendicular to the image G in the up-down direction. The image rotation unit 14 rotates the image G clockwise or counterclockwise at every fixed angle (for example, 5 ° or 10 °). Then, while the image rotation unit 14 rotates the image G, the connection ratio calculation unit 13 described above calculates, for each rotation angle with respect to the vertical direction, the degree to which the extending direction of the crack pixel P1 coincides with the vertical direction by the integration value described above. Is calculated as

  FIG. 9B illustrates an example in which the image rotation unit 14 does not rotate the image G, FIG. 9A illustrates an example in which the image rotation unit 14 rotates the image G by 50 degrees clockwise, and FIG. An example in which the image rotation unit 14 rotates the image G clockwise by 20 ° is shown. In these examples, the integration value (degree) calculated by the connection ratio calculation unit 13 is small when the image G is rotated clockwise by 50 ° (FIG. 9A), and when the image G is not rotated ( 9B, and becomes the largest when the image G is rotated by 20 degrees clockwise (FIG. 9C).

  The angle detection unit 15 (determination unit) detects the direction in which the crack in the image G extends from the integrated value for each rotation angle described above. The angle detection unit 15 determines that the direction in which the calculated integral value is rotated from the vertical direction in the opposite direction by the highest rotation angle is the direction in which the crack extends. In the example illustrated in FIGS. 9A to 9C, the angle detection unit 15 determines that the direction rotated counterclockwise from the vertical direction by 20 ° is the direction in which the crack extends.

  The interval calculation unit 16 calculates an interval between the cracks after the angle detection unit 15 determines the direction in which the crack extends. For example, when the integrated value (degree) is maximized when the image G is rotated clockwise by 20 ° as shown in FIG. 9C, the image G is rotated clockwise by 20 ° in the vertical direction. Calculate the distance between cracks in this state.

  Specifically, as shown in the example of FIG. 10, in the image G rotated clockwise by 20 °, the interval calculation unit 16 moves on the virtual line L extending in the left-right direction (vertical direction perpendicular to the vertical direction). By dividing the total number of located pixels (pixels) by the number of crack pixels P1 located on the imaginary line L, a crack interval is calculated.

  In the example shown in FIG. 10, the interval calculation unit 16 calculates the number of crack pixels P1 for each virtual line L extending in the left-right direction. The numerical value on the right side of the image G in FIG. 10 indicates the calculated number of crack pixels P1. As described above, after calculating the number of crack pixels P1 for all the virtual lines L, the interval calculation unit 16 calculates the average value (or the maximum value) of the number of crack pixels P1.

  In the example of FIG. 10, the interval calculation unit 16 sets the total number of pixels located on the virtual line L extending in the left-right direction to 20, and sets the average value of the crack pixels P1 located on the virtual line L to 2.6. “7.7” obtained by dividing 20 by 2.6 is calculated as the crack interval. When dividing 20 by the maximum value of the crack pixel P1 located on the virtual line L, the interval calculation unit 16 calculates “5.0” obtained by dividing 20 by 4 as the crack interval.

  The display control unit 17 displays on the monitor 3 the direction in which the crack determined by the angle detection unit 15 extends. The display control unit 17 displays the direction in which the crack extends for each region R (image G) extracted by the image extraction unit 12. Specifically, as shown in FIG. 11, when each region R has a square shape, the display control unit 17 indicates a direction in which a crack extends in each region R as a line segment N.

  The direction in which the line segment N extends indicates the dominant direction of the crack determined by the angle detection unit 15, and the length of the line segment N indicates the above-described integral value (degree). That is, the longer the length of the line segment N, the stronger the continuity of the crack in the predominant direction.

  In addition, the display control unit 17 determines that the line segment N (crack is not formed) only in the region R in which the integral value (degree) exceeds the first threshold value and the interval calculated by the interval calculation unit 16 is equal to or less than the second threshold value. Direction). In FIG. 11, a region R in which the line segment N is not displayed is a region R in which the integral value (degree) is equal to or less than the first threshold value or in which the interval calculated by the interval calculation unit 16 is larger than the second threshold value. It is.

  The above-described first threshold value indicates a threshold value of the strength of continuity of the crack that indicates the direction in which the crack extends, and the second threshold value indicates a threshold value of the interval (density) of the cracks that indicates the direction in which the crack extends. ing. The first threshold value and the second threshold value are set values and can be changed as appropriate.

  After displaying the line segment N as described above, the display control unit 17 erases the image of the tunnel face T and leaves only the outer edge of the analysis area D and the line segment N as shown in FIG. State. Then, as shown in FIG. 13, the display control unit 17 compares the continuity of cracks and the dominant direction between a plurality of regions R adjacent to each other in the line segment N. After that, the display control unit 17 smoothly connects the plurality of line segments N having high continuity and the plurality of line segments N having similar prominent directions with the curve C.

  After generating the curve C as described above, the display control unit 17 deletes the line segment N and leaves only the outer edge of the analysis area D and the curve C as shown in FIG. A sketch image of a crack in the tunnel face T is obtained. As described above, in the crack analysis system 1, since the sketch image of the crack of the tunnel face T is obtained by the display control unit 17, it is possible to save the labor of drawing the sketch by the site worker.

  In addition, the sketch image generated by the display control unit 17 is not displayed manually by the site worker, but is automatically displayed, so that the quality of the sketch of the tunnel face T does not vary. Therefore, it is possible to eliminate the variation in accuracy that may occur in grasping the state of the crack in the tunnel face T.

  Next, a crack analysis method according to the present embodiment will be described. FIG. 15 is a flowchart illustrating an example of a crack analysis method using the crack analysis system 1. Hereinafter, an example of analyzing a crack generated in the tunnel face T which is the inspection object illustrated in FIG. 2 will be described.

  First, the crack analysis system 1 obtains an image of the tunnel face T by the image obtaining unit 2. At this time, the image acquisition unit 2 acquires digital data of the image of the tunnel face T, and the image processing unit 10 receives the digital data acquired by the image acquisition unit 2 (Step S1).

  Next, the image binarization unit 11 binarizes the image of the tunnel face T into a crack pixel P1 and a non-crack pixel P2 (first step). Specifically, the image binarization unit 11 creates an analysis area D for the received digital data, and performs monochrome binarization on an image inside the analysis area D (Step S2).

  Subsequently, the image extraction unit 12 divides the analysis region D generated by the image binarization unit 11 into a plurality of regions R. Then, the image extracting unit 12 extracts an arbitrary portion from the plurality of divided regions R, that is, extracts the region R of the tunnel face T (step S3). The connection rate calculation unit 13 calculates the degree to which the extending direction of the crack pixel P1 coincides with the region R extracted by the image extraction unit 12 in the up-down direction. At this time, the connection ratio calculation unit 13 divides the image G into a plurality of divided images B, calculates the connection ratio, calculates the relational expression between the size of the divided image B and the connection ratio, and integrates the relational expression (degree ) Is calculated (step S4).

  The above-described step S4 is executed while rotating the image G by the image rotation unit 14 by 180 ° (90 ° clockwise and 90 ° counterclockwise) with respect to the vertical direction. Specifically, the image G is rotated by a fixed angle clockwise from the vertical direction (reference direction) to calculate the integral value, and after the image G is rotated 90 ° to calculate the integral value, the image G is moved up and down. The integral value is calculated while rotating counterclockwise by a fixed angle from the direction, and after calculating the integral value by rotating counterclockwise by 90 °, all angles are ended.

  Therefore, the process proceeds from step S4 to step S5, and if it is determined in step S5 that all the angles have not been completed (NO in step S5), the process proceeds to step S6, and the image rotating unit 14 Rotate a fixed angle. Then, the process proceeds to step S4, where the integral value (degree) is calculated again (second step).

  If it is determined in step S5 that all the angles have been completed (YES in step S5), the angle detection unit 15 determines the crack of the image G from the plurality of integrated values obtained through the plurality of steps S4. The direction in which is extending is detected. At this time, the angle detection unit 15 determines that the direction rotated from the vertical direction to the opposite direction by the rotation angle at which the integral value obtained in step S4 is the highest is the direction in which the crack is extending (step S7, step S7). 3 steps).

  Next, the interval calculation unit 16 calculates the interval between the cracks. As described above, the interval calculation unit 16 divides the total number of pixels located on the virtual line L in the image G of the region R by the number of crack pixels P1 located on the virtual line L, thereby An interval is calculated (step S8).

  Subsequently, the process proceeds to step S9, and it is determined whether or not all the regions R have been completed, that is, whether or not the steps S4 to S8 have been completed for all the regions R. If it is determined in step S9 that all the regions R have not been completed (NO in step S9), the process returns to step S3, where the image extracting unit 12 extracts another region R, and performs a series of steps S4 to S8. Perform the process.

  On the other hand, when it is determined in step S9 that all the regions R have ended (YES in step S9), the display control unit 17 performs display control of a crack image to be displayed on the monitor 3. At this time, as shown in FIGS. 11 to 14, the display control unit 17 shows the line segment N for each region R, and leaves only the outer edge of the analysis region D and the line segment N to form a plurality of lines with the curve C. The segment N is smoothly connected to eliminate the line segment N, thereby generating a sketch image of the crack (step S10). Then, the display controller 17 displays the generated sketch image on the monitor 3 (step S11, fourth step), and a series of steps is completed.

  Next, effects obtained from the crack analysis method and the crack analysis system 1 according to the present embodiment will be described.

  In the crack analysis method according to the present embodiment, as shown in FIG. 9, the image G is rotated, and the degree of coincidence of the extension direction of the crack pixel P1 with respect to the vertical direction is calculated for each rotation angle with respect to the vertical direction. . Then, a rotation angle having a high degree of coincidence is specified, a direction rotated in the opposite direction from the vertical direction by the specified rotation angle is determined as a direction in which the crack extends, and the direction is displayed. Therefore, by performing the image processing for rotating the acquired image G, the direction in which the crack extends in the tunnel face T can be grasped.

  Therefore, it is possible to output data that can easily grasp the state of occurrence of cracks to a site worker who manages the tunnel face T. As described above, the direction in which the crack extends at the tunnel face T is displayed, so that the field worker can intuitively grasp the state of the crack occurrence.

  The second step described above includes, as shown in FIGS. 5 to 7, a step of dividing the image G into a plurality of divided images B, and a step of dividing the crack pixel P1 from the upper end to the lower end with respect to the total number of the divided images B. Calculating the ratio (connection rate) of the number of extended divided images B. Therefore, the degree of coincidence of the extending direction of the crack pixel P1 with respect to the vertical direction can be calculated as a numerical value.

  Further, since the ratio (connection ratio) calculated as described above is a ratio of the number of divided images B in which the crack pixels P1 extend from the upper end to the lower end, a simple algorithm compared with a predetermined calculation such as a fractal dimension is used. It is. Therefore, the calculation process in each step can be speeded up and the data output can be speeded up, so that the data indicating the state of occurrence of cracks can be quickly output to the site worker.

  Further, the above-described dividing step and the step of calculating the ratio (connection rate) are performed a plurality of times while changing the number of the divided images B, and as shown in FIG. And a step of calculating an integral value of the relational expression as the above-described degree. Therefore, by calculating the degree to which the extension direction of the crack pixel P1 coincides with the vertical direction as an integral value, the direction in which the crack extends can be determined with higher accuracy. Therefore, the extending direction of the crack to be displayed can be output as data that matches the actual feeling of the site worker.

  As shown in FIGS. 9C and 10, after the third step described above, the image G is rotated in the vertical direction by the rotation angle having the highest calculated integral value, that is, the most. In a state where the crack pixel P1 extends in the vertical direction, the total number of pixels located on the virtual line L extending in the horizontal direction is divided by the number of crack pixels P1 located on the virtual line L, thereby Calculating an interval. As described above, by calculating the interval between the cracks, the density of the cracks generated in the tunnel face T can be grasped. Therefore, the state of occurrence of cracks can be grasped with higher accuracy.

  As shown in FIG. 11, the above-described second step, third step, and fourth step are performed for each of the plurality of regions R of the tunnel face T, and in the fourth step, the highest value is set in the third step. The directions in which the cracks extend are shown only in the region R where the integral value exceeds the first threshold. Therefore, the direction in which the crack extends is displayed only in the region R in which the crack pixels P1 extend in a certain direction. Therefore, by selectively extracting cracks that need to be grasped, the field worker can grasp the cracks more reliably.

  The second step, the third step, the fourth step, and the step of calculating the interval are performed for each of the plurality of regions R of the tunnel face T. In the fourth step, the interval calculated in the step of calculating the interval is The directions in which the cracks extend are displayed only for the area equal to or less than the second threshold. Therefore, the direction in which the crack extends is displayed only in the region R where the gap is narrow and the crack is formed more densely. Therefore, by selectively extracting cracks that need to be grasped, the field worker can grasp the cracks more reliably.

  Further, in the crack analysis system 1 according to the present embodiment, as shown in FIG. 9, the image rotation unit 14 rotates the image G, and the connection ratio calculation unit 13 determines the direction in which the crack pixel P1 extends in the vertical direction. Calculate the degree of coincidence. The angle detection unit 15 specifies a rotation angle having a high degree of coincidence, and determines a direction in which the rotation is performed in the opposite direction from the vertical direction by the specified rotation angle as a direction in which the crack extends. Then, the monitor 3 displays the direction in which the determined crack extends. Therefore, the image rotating unit 14 rotates the image G, and the angle detecting unit 15 determines the direction in which the crack extends, whereby the extending direction of the crack generated in the tunnel face T can be grasped. Therefore, the same effect as the above-described crack analysis method can be obtained.

  As described above, the embodiments of the present invention have been described. However, the present invention is not limited to the above-described embodiments, and may be modified or applied to other matters without changing the gist described in each claim. It may be something. That is, the present invention can be variously modified without changing the gist of each claim.

  For example, in the above-described embodiment, an example has been described in which the degree of coincidence of the extension direction of the crack pixel P1 in the up-down direction is calculated by the connection rate calculation unit 13 calculating the connection rate. However, the method of calculating the degree to which the extending direction of the crack pixel P1 coincides with the vertical direction may be executed by means other than the connection ratio. Further, the connection ratio calculation unit 13 does not need to calculate the above-described integral value. Further, in the above-described embodiment, an example in which the reference direction is the vertical direction has been described, but a direction other than the vertical direction may be the reference direction.

  Further, in the above-described embodiment, an example has been described in which the analysis region D has an arc shape including the lower left portion A1, the lower right portion A2, and the top portion A3. However, the shape of the analysis region D is not limited to the arc shape and can be changed as appropriate. It is.

  In the above-described embodiment, the display controller 17 generates the line segment N and generates the curve C connecting the plurality of line segments N, thereby displaying the sketch image of the crack of the tunnel face T on the monitor 3. Examples have been described. However, the method of displaying the direction in which the crack extends is not limited to the above example, and can be appropriately changed. For example, the direction in which the crack extends, the continuity of the crack, and the interval between the cracks may be displayed on the monitor 3 by numerical values.

  In the above-described embodiment, an example in which the inspection object of the crack analysis system 1 and the crack analysis method is the tunnel face T has been described. However, the crack analysis system 1 and the crack analysis method can be applied to other rocks such as a dam embankment, and can also be applied to non-rocks such as concrete.

DESCRIPTION OF SYMBOLS 1 ... crack analysis system, 2 ... image acquisition part, 3 ... monitor (display part), 10 ... image processing part, 11 ... image binarization part, 12 ... image extraction part, 13 ... connection rate calculation part (degree calculation part) ), 14: image rotating unit (degree calculating unit), 15: angle detecting unit, 16: interval calculating unit, 17: display control unit, A1: lower left, A2: lower right, A3: top, B: divided image, C: Curve, D: Analysis area, G: Image, L: Virtual line, N: Line segment, P1: Cracked pixel, P2: Non-cracked pixel, R: Area, S: Area, T: Tunnel face (object to be inspected) ).

Claims (6)

A crack analysis method for analyzing a crack occurring in an inspection object,
Obtain an image of the inspection object, the image, the first step of binarizing the image into crack pixels indicating the portion where the crack has occurred, and non-crack pixels indicating the portion where the crack has not occurred.
While rotating the image about an axis perpendicular to the image with respect to a reference direction, calculating a degree at which the extension direction of the crack pixel matches the reference direction for each rotation angle with respect to the reference direction. Steps and
A third step of determining the direction in which the calculated degree is rotated by the highest rotation angle in the opposite direction from the reference direction as the direction in which the crack extends,
A fourth step of displaying a direction in which the determined crack extends;
Equipped with a,
The second step includes:
Dividing the image into a plurality of divided images;
Calculating the ratio of the number of the divided images in which the crack pixels extend from one end to the other end in the reference direction with respect to the total number of the divided images;
A crack analysis method having the following . The step of dividing and the step of calculating the ratio are performed a plurality of times while changing the number of the divided images,
Calculating a relational expression between the size of the divided image and the ratio,
Calculating the integral value of the relational expression as the degree,
The crack analysis method according to claim 1 , further comprising: After the third step,
In a state where the image is rotated by the rotation angle with respect to the reference direction, the total number of the pixels located on a virtual line extending in a direction orthogonal to the reference direction is calculated by the number of the crack pixels located on the virtual line. Further comprising the step of calculating an interval between the cracks,
Crack analysis method according to claim 1 or 2. The second step, the third step, and the fourth step are performed for each of a plurality of regions of the inspection object,
In the fourth step, a direction in which the crack extends is displayed only in the region in which the degree determined as the highest in the third step exceeds a first threshold value,
Crack analysis method according to any one of claims 1-3. The second step, the third step, the fourth step, and the step of calculating the interval is performed for each of a plurality of regions of the inspection object,
In the fourth step, the direction in which the crack extends is displayed only for the region in which the interval calculated in the step of calculating the interval is equal to or less than a second threshold value.
The crack analysis method according to claim 3 . A crack analysis system for analyzing a crack generated in an inspection object,
An image binarization unit that acquires an image of the inspection object and binarizes the image into a crack pixel indicating a portion where the crack has occurred and a non-crack pixel indicating a portion where the crack has not occurred. When,
While rotating the image about an axis perpendicular to the image with respect to the reference direction, for each rotation angle with respect to the reference direction, calculate a degree at which the extension direction of the crack pixel matches the reference direction. Department and
A determination unit that determines the direction in which the calculated degree is rotated in the opposite direction from the reference direction by the highest rotation angle, as the direction in which the crack extends,
A display unit that displays the direction in which the determined crack extends,
Equipped with a,
The degree calculating unit is a connection ratio calculating unit that divides the image into a plurality of divided images,
The connection ratio calculation unit calculates a ratio of the number of the divided images in which the crack pixels extend from one end to the other end in the reference direction with respect to the total number of the divided images.
Crack analysis system.