JP4372862B2 - Concrete deterioration degree quantification method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a concrete deterioration degree quantifying method for inspecting a deterioration state of concrete used for a floor slab of a steel bridge or the like.
[0002]
[Prior art]
Conventionally, the deterioration state of concrete used for floor slabs such as steel bridges has been visually diagnosed by a specialized inspector, and thus cannot be evaluated objectively.
[0003]
In view of this, an apparatus has been proposed (Japanese Patent Laid-Open No. 6-229930) that processes a picture of a concrete surface to extract a cracked portion and calculate and quantify the feature amount.
[0004]
[Problems to be solved by the invention]
However, many of the conventional concrete deterioration degree quantifying apparatuses cannot use ordinary cameras for photographing concrete surfaces, and require infrared cameras. In addition, there are devices that can use images taken with a normal camera. However, when inspecting a concrete bridge slab of a steel bridge, there is a practical problem because it is necessary to assemble a scaffold and perform close-up photography.
[0005]
In addition, other than the object, for example, members such as steel materials and pipes are often reflected at the time of shooting, so in order to obtain a reliable extraction result, these must be excluded manually. In other words, it is difficult to automate the processing.
[0006]
Then, the objective of this invention is providing the concrete degradation degree quantification method which can extract the crack and a free lime from the image containing an unnecessary member which solves the said subject.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, the invention of claim 1 takes in the image information of the concrete surface to be inspected, and uses the difference in density at the boundary between the concrete surface area in the original image and the area of unnecessary members such as piping. The edge candidates are extracted, the edge candidates are described in segments, the segments on the straight line are connected, and the edges are joined together to create a closed area surrounded by the joined edges in the image. It is determined whether it is a concrete surface using the geometrical characteristics and density statistics of the region, and a mask image is formed with the unnecessary member region as the mask region, and the concrete surface region is determined from the original image using the mask image. It is a concrete deterioration degree quantification method in which after extraction, cracks and free lime are extracted from the concrete surface area and their feature values are calculated.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.
[0010]
First, a system configuration for carrying out the concrete deterioration degree quantification method will be described with reference to FIG.
[0011]
In FIG. 2, 10 is a video deck, 11 is an image input device such as a camera, 12 is an image memory (frame memory) 13 for storing an image input from the video deck 10 or the image input device 11, and An image monitor for displaying an input image, 14 is an engineering workstation, 15 is a console display, 16 is a touch panel provided in front of the console display 15, and 17 is an image processing apparatus.
[0012]
In order to obtain image information of a concrete surface to be inspected, such as a concrete floor slab, when the concrete surface is photographed with a video camera, from the video deck 10, and when photographed with a camera, the image memory 12 from the image input device 11. The degraded portion is extracted by the image processing device 17 from the captured image, and the processing result image has the feature amount such as the area ratio and the length of the degraded portion. At the same time, it is displayed on the console display 15, and at the same time, the input image is displayed on the image monitor 13.
[0013]
Image processing operations can be easily performed by touching operation buttons displayed on the screen of the display 15 using the touch panel 16 attached to the front surface of the console display 15.
[0014]
In extracting the deteriorated part of the concrete, there are firstly unnecessary parts such as pipes and steel on the concrete surface, and these are not subject to inspection in the image, and cracking and free lime are generated as a concrete deterioration phenomenon. There are two, which need to be checked separately.
[0015]
The generation of this free lime appears as a phenomenon in which rainwater infiltrates into the generated crack, the lime content of the concrete oozes out from the crack and the surface becomes white, and it is different from the lightness of the crack. Need to be inspected.
[0016]
A flow for extracting the deteriorated portion will be described with reference to FIG.
[0017]
Control is started 20, the image is input 21, and in step 1, if there is an unnecessary part such as a water pipe in the image (Yes), after removing the water pipe etc. 22 from the image, step 2 If not (no), move to step2.
[0018]
In step 2, it is determined whether or not to extract a crack. If there is a crack in the image (yes), after extracting the crack 23, the process proceeds to step 3, and if not (no), the process proceeds to step 3.
[0019]
In step 3, it is determined whether or not free lime is extracted. If there is free lime in the image (yes), free lime extraction 24 is performed and result display 25 is performed to finish the control 26. (No) The result display 25 is performed without performing the extraction 24 of free lime, and the control is terminated 26.
[0020]
Next, each process of FIG. 1 will be described.
[0021]
(1) Removal of unnecessary parts (distribution pipes, etc.) If the input image contains piping other than the target concrete surface, the difference between the unnecessary parts and the boundary with the concrete surface In addition, the surface of the concrete and the unnecessary member are automatically divided into regions by image processing using the feature that the boundary forms a closed loop.
[0022]
(2) Extraction of free lime Free lime has a feature that there is a very bright (white) region at the center and a lighter white region spreads around it. Using this feature, first, by image processing, that is, binarizing the image with a high threshold to extract a core region, and then binarizing sequentially with a threshold slightly lower than the high threshold, a thin white around the nucleus Extract by expanding and consolidating areas. In addition, free lime that has turned black due to rust flowing out of the reinforcing bars is extracted using the characteristics that are black compared to normal concrete surface and adjacent to normal (white) free lime.
[0023]
(3) Extraction of cracks Crack candidates are extracted by utilizing the normal concrete surface and the difference in density between cracks. Since these candidates include mis-extracted ones, attention is focused on continuity from among the crack candidates, and the mis-extraction candidates are removed and the candidates are combined to extract the cracks.
[0024]
(4) Removal of formwork traces Formwork traces may be generated on the concrete surface, like a concrete floor slab of a road bridge. Since the formwork traces appear black as compared to the normal concrete surface as well as cracks, they may be erroneously extracted as cracks. Therefore, the mold trace is removed from the crack extraction result by utilizing the feature of the mold trace that is long and strong in linearity compared with the crack.
[0025]
(5) Feature amount calculation From the extraction result of free lime, the number and area ratio of free lime are calculated from the extraction result of cracks, and the number and length of cracks are calculated and displayed on the console display 15.
[0026]
Next, the unnecessary member removal method (1), which is a feature of the present invention, will be described in detail with reference to FIGS.
[0027]
As shown in FIG. 3, in the flow of removing unnecessary members, removal control is started 30, an edge candidate extraction process 31 is performed, a segment description 32 is performed on the extracted edge candidates, and then a global search by hypothesis verification is performed. 33, a region dividing process 34 is performed to create a closed region, an unnecessary member removing process 35 is performed based on this, and the control is terminated 36.
[0028]
These processes will be described below with reference to FIG.
[0029]
FIG. 4A shows the input original image 40, in which an area 41 of unnecessary members such as steel is reflected, and the rest is a concrete surface area 42, and in the concrete surface area 42. Suppose that the crack 43 and the free lime 44 are reflected.
[0030]
(1) Unnecessary Member Edge Candidate Extraction Processing The unnecessary member edge candidate extraction is performed by using an edge extraction process or the like using a portion where the density is rapidly changed as a boundary portion as shown in FIG. The edge candidate extraction processing image 45 is obtained.
[0031]
(2) Segment Description A continuous region of edge candidates 46 in the image 45 in FIG. 4B is regarded as a straight line as a segment, and a segment 48 is described in the image 47 as shown in FIG. 4C. A segment estimated as a part of a true member edge from the group of segments 48 is set as a search starting point.
[0032]
(3) Global search processing by hypothesis verification The segment constituting the edge of the unnecessary member is verified by the hypothesis that the edge is distributed in a straight line, so that the segment 48 of the image 47 in FIG. As shown in FIG. 4D, the segments 48 are connected to form an image 50 in which the edge 49 is searched globally.
[0033]
(4) Region splitting processing of the concrete floor slab and unnecessary members The edges of the unnecessary members are joined to the edges 49 in FIG. A closed region 53 surrounded by a combined edge 52 is created in the image 51 shown.
[0034]
(5) Unnecessary member removal processing It is determined whether each region 53 divided by the above-described processing is an unnecessary member or a concrete floor slab, and a mask image 55 is created by masking the unnecessary member region 54 as shown in FIG. To do. The determination as to whether it is an unnecessary member or a concrete slab is made using the geometrical characteristics of the region and the concentration statistics.
[0035]
Thus, using the mask image 55 created by the processes (1) to (5), the concrete surface area 42 in the original image 40 shown in FIG. Free lime 44 is extracted.
[0036]
If the cracks 43 and the free lime 44 are extracted in this way, the number of cracks, the length of each crack, the number of free limes and the area ratio thereof are calculated, and the feature quantities of the cracks and free lime are quantified. Therefore, it is possible to evaluate the degree of deterioration with higher accuracy.
[0037]
【The invention's effect】
In short, according to the present invention, even if an unnecessary member such as a pipe is shown in the image, a mask image is created based on the unnecessary member and only the concrete surface area is extracted, thereby accurately detecting cracks and free lime. It is possible to extract well and evaluate the degree of deterioration.
[Brief description of the drawings]
FIG. 1 is a diagram showing a flow of image processing according to the present invention.
FIG. 2 is a diagram illustrating an outline of a system configuration in the present invention.
FIG. 3 is a diagram showing a flow of extracting unnecessary members in the present invention.
FIG. 4 is a diagram illustrating a process of removing unnecessary members from an original image in the present invention.
[Explanation of symbols]
40 Original images 41 and 54 Unnecessary member region 42 Concrete surface region 43 Crack 44 Free lime 46 Edge candidate 55 Mask image

Claims (1)

The image information of the concrete surface to be inspected is taken in, the edge candidate is extracted by using the gray level difference between the concrete surface area in the original image and the area of unnecessary parts such as pipes, and the edge candidate is described in segments. In addition, the segments on the straight line are connected to join the edges together, creating a closed region surrounded by the combined edges in the image, and whether the closed region is an unwanted member or a concrete surface. Judgment is made using the amount, and a mask image is formed using the unnecessary member area as a mask area. After the concrete surface area is extracted from the original image using the mask image, cracks and free lime are extracted from the concrete surface area. And a concrete deterioration degree quantification method characterized by calculating these feature amounts.

Images