JP2541735B2 - Method and apparatus for diagnosing coating film deterioration - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for inspecting deterioration due to rust or peeling of a coating film on a steel bridge or the like, and in particular, quantitatively detecting deterioration from partial characteristics while quantitatively detecting deterioration from overall characteristics. However, the present invention relates to a coating film deterioration diagnosing system which judges the degree of deterioration progress quantitatively without being influenced by the partial state of rust or peeling by comprehensively judging these detection results.

[0002]

2. Description of the Related Art A steel bridge constructed mainly of steel outdoors and having a protective coating formed on its surface,
The coating film deteriorates over time. The deterioration of the coating film includes cracking, peeling, and leaching of rust on the steel material. If the steel material or rust is exposed due to such deterioration of the coating film, the deterioration of the steel material itself is accelerated. Further, the deterioration of the coating film impairs the aesthetics of the steel bridge. Such deterioration of the coating film gradually progresses over a long time. In order to detect deterioration of the coating film, regular inspection is performed after a certain period of time. If there is deterioration of the coating film as a result of the inspection, it is necessary to carry out repair coating to recover it. If the repair coating is performed early, the coating will always be in good condition.
On the other hand, repair painting costs a lot of money, so we want to reduce the number of times as much as possible. If repair painting is performed when the deterioration is not so great, the number of times increases and the cost increases. Therefore, it is important to judge at which point (how much the coating film has deteriorated) to perform the repair coating.

A method for inspecting deterioration of a coating film on a steel bridge or the like and a method for deciding whether to carry out repair coating, which are generally used in the past, will be described.

In the regular inspection, first, an inspector goes to the site, takes a photo or the like on the spot or takes it home, and determines the degree of deterioration. The basis of the judgment is the number of cracks, peeling, leaching of rust of the steel material (hereinafter collectively referred to as a deformed portion), the area, or the like. If the number of deformed portions is large or the area is large, it is determined that the deterioration is progressing. In order to make this judgment as objectively as possible, it is judged by the scoring method. In the scoring method, for example, the best condition is 100 points, and 60 points according to the progress of deterioration.
It is a method of scoring 40 points. A limit standard is set in advance for the points scored in this way, and if it is less than that score, it is decided to carry out repair painting.

[0005]

Since the conventional inspection for deterioration of coating film relies on the judgment of a person, individual differences due to experience or the like will appear. For example, the size of the deformed portion ranges from minute to huge, and the shape is also indefinite. Even in the above method, how many deformed portions are recognized as deformed portions varies from person to person, so the number and area are not constant. Therefore, the scoring for the same inspection target differs depending on the person.

Further, since many steel bridges and the like may be provided along one road, it is preferable to carry out repair painting systematically. However, in order to make a plan, it is necessary to know in advance which steel bridge will reach the limit criterion of deterioration. In the conventional method, the determination of deterioration during inspection is ambiguous, and it is difficult to predict the progress of deterioration after inspection.

In order to solve such a problem, a means for quantitatively detecting the presence or absence and the size of the deformed portion is required.
To this end, it is preferable to introduce a generally known image processing technique. For example, it is possible to quantitatively detect the presence or absence and the size of a deformed portion by capturing an image of an object to be inspected as grayscale image information and detecting a local change in grayscale as a change in the coating film.

Generally, when an object to be inspected is imaged as grayscale image information, the amount of reflected light is affected when the light from the light source when the grayscale image is taken is not uniformly applied to the object to be measured. Shading appears in the grayscale image. When the shading is a relatively gradual change in shade based on the light source, the shading can be removed by performing image processing so as to leave only a relatively rapid change in shade of the deformed portion. However, since steel bridges and the like are outdoors and there is a three-dimensional layout of the structures, it is inevitable that the shadows of other structures will appear in the image. in this case,
Shading is very strong. It has been difficult in the prior art to remove the shadows of other objects that show such a rapid change in shading. Also, when the deformed portion is evenly distributed over a wide area of the screen, the change in shading becomes monotonous, so just by examining the local shade change of the image as described above, Will not be detected.

Therefore, an object of the present invention is to solve the above-mentioned problems, to quantitatively detect deterioration from the characteristics of a part, and quantitatively to detect the characteristics of the whole, and to make a comprehensive determination of these detection results. It is an object of the present invention to provide a coating film deterioration diagnosing method and apparatus that quantitatively judge the degree of deterioration progress regardless of the partial state of rust or peeling.

[0010]

[Means for Solving the Problems] In order to achieve the above object, the present invention is directed to cracking deterioration of a coating film on a steel building such as a steel bridge ,
Peeling off, in order to check on the basis of the variable-shaped part consisting of leaching steel rust, imaging the inspection object as a gray-scale image, the dark
For coating film deterioration diagnosis method that detects a deformed portion from light image information
The gray value changes from dark or light to dark or light from the above gray image information.
The deviation until the return to the original is within the predetermined threshold.
The filtered image formed by averaging the gray values of the
The part of the image that is subtracted from the original image and that is greater than or equal to the specified gray level threshold
Minute is detected as a partial deformed portion and
Area of each partial deformation part, partial deformation over a certain area
The number of parts and the area of the entire image
The ratio of the area occupied by the area is calculated, and the grayscale image information is divided into small areas at predetermined intervals, and a statistic consisting of the average value and the variance of the gray values included in the small areas is calculated for each small area.
Average be smaller dispersion and dispersion of larger small region from this statistic is determined to statistically Henjo portion and a greatly different small regions from the average value of the entire image, the number of the statistical Henjo portion predetermined If the value is equal to or larger than the value, it is determined that there is deterioration, and on the other hand, from the deformed portion data, there is a partial deformed portion having an area equal to or larger than a predetermined value, or there is a predetermined number or more of partial deformed portions having an area equal to or larger than a predetermined value. Or, the surface of the partial deformation part with respect to the area of the entire image
When the ratio of the product exceeds a predetermined value, it is determined that there is deterioration, and the deterioration determination based on these statistics and the deterioration determination based on the deformed portion data are combined to determine the degree of deterioration progress of the inspection target. It was done.

[0011]

Prior to the present invention, the present inventors have announced the following grayscale morphology processing. The shading Moforojii process, there is a bias in the gray value from the gray-scale image
When it exists over a wide area within the range, that portion is detected as a significant gradation change portion. In the light and shade morphology processing, the width from when the change of light and shade appears on the dark side or the light side until it returns to the original shade by the opposite change,
That is, a threshold is provided in advance for the width of a portion that is biased to one of the shades, and the biased portion of the shade within the threshold is only on the dark side.
Alternatively, only the light side is flattened once, and this flattening is performed on the entire image.
The image applied to is used as a filtered image. These filtered images are subtracted from the original image. This allows shading to be removed from the original image. From the image after shading removal, a portion equal to or larger than a predetermined grayscale threshold value is detected as a deformed portion. By doing so, a significant shade change portion can be detected as a deformed portion of the coating film showing rust, peeling, or the like. The deformed portion detected in this way is
In order to distinguish it from the statistical deformation part described later,
I will call it a deformed part.

With respect to the partial deformed portion thus detected , the area of each partial deformed portion, which is equal to or larger than a predetermined area,
The number of deformed parts and the partial deformed part with respect to the area of the entire image
The ratio occupied by the entire area is obtained and used as the deformed portion data.
The deformed portion data is obtained by quantitatively detecting partial characteristics of the image.

Next, the grayscale image information obtained by normalizing the original image is divided into small regions at a predetermined interval, and a statistical amount consisting of an average value and a variance of the grayscale values contained in the small regions is obtained for each small region. Further, the average value of these small areas and the average value and dispersion of the entire image are calculated. Statistics in a small area, concentrated
The following tendencies are shown depending on the presence or absence of light bias and the size.

(1) The variance is close to 0 when the image is uniform with no unevenness in shade .

(2) The dispersion is large when the lightness changes due to the presence of the lightness unevenness .

(3) When there is a large unevenness of shading ,
Although the variance is small in that small region, the average value is different from the average values in other regions. The average value of the small area has a large difference from the average value of the whole.

The statistics of each in this manner small area is, of light and shade
The presence or absence of bias, exhibits different trends by size, chromatic
A deformed portion (hereinafter referred to as
This will be referred to as a statistically deformed portion) . The statistic is a quantitative detection of the overall characteristics of the image.

Therefore, the degree of progress of deterioration is determined by using the detection result of the statistical deformed portion based on this statistic and the detection result of the partial deformed portion based on the deformed portion data. This makes it possible to quantitatively determine the degree of deterioration progress without being affected by the partial state of rust or peeling.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As shown in FIG. 1, a coating film deterioration diagnosis system 1 for inspecting the deterioration of a coating film on a steel bridge or the like has a CCD camera 3 provided facing an object 2 to be inspected, and a CCD camera 3.
The deformed part extraction means 4 and the statistical means 5 which receive the image output of the above, and the deterioration degree determination means 6 which receives the outputs from the deformed part extraction means 4 and the statistical means 5 are provided.

The CCD camera 3 is an image pickup means 7 capable of picking up an image of the object 2 to be inspected as grayscale image information.
The CCD camera 3 has pixels of the order of a normal television image, and each pixel has a grayscale resolution of about 256 gradations.

The deformed portion extracting means 4 is present over a range within a range in which the gray value is biased from the gray image information captured by the CCD camera 3 based on the gray morphology processing.
When that part is present, that part is significantly unevenly shaded, that is, rusted or peeled.
It can be detected as a partially deformed portion indicating such as . In addition, the area of each partially deformed portion, a portion of a predetermined area or more
Partial deformation with respect to the number of dynamic deformation parts and the area of the entire image
Calculate the ratio of the area of the entire part ,
It is configured to send the deformed portion data including the number and the ratio to the deterioration degree determination unit 6.

The statistical means 5 normalizes the grayscale image information picked up by the CCD camera 3 and divides the grayscale image information into small areas at predetermined intervals, and also comprises, for each small area, an average value and a variance of the grayscale values contained in the small area. The statistic amount can be obtained and sent to the deterioration degree determination means 6.

The deterioration degree judging means 6 can judge the degree of progress of deterioration of the object to be inspected from the statistic and the deformed portion data in accordance with a predetermined judgment standard and rule. In addition, the deterioration degree determining unit 6 can output a subsequent statistical amount or a predicted value of the deformed portion data based on the statistical amount or the deformed portion data of the inspection target 2 obtained over time. .

FIG. 1 shows a control device 8 incorporating the above-mentioned deformed portion extracting means 4, statistical means 5 and deterioration degree determining means 6. The control device 8 includes a CCD camera 3 and a VTR.
Device 9, communication modem 10 that can use communication line, judgment database 11 that stores judgment criteria and rules and provides statistical means 51, history database 12 that accumulates judgment results for each inspected object 2, display device 13 such as CRT, control A mouse 14 and a keyboard 15 for operating the device 8 and a printer 16 for printing out the determination result and the like are connected. The control device 8 also contains an image memory 17. The image memory 17 stores the grayscale image information from the CCD camera 3,
The video image from the VTR device 9 can be converted into grayscale image information and stored. The display 13 can display image information from the CCD camera 3 or the VTR device 9, image information processed by the deformed portion extracting means 4 and the statistical means 5, determination results, and the like.

Next, the operation of the embodiment will be described.

A predetermined position on the coating film surface of a steel bridge or the like is set as the object 2 to be inspected. After that, this inspection position is stored in the history database 12
Is fixed and the camera angle is specified, but the present invention allows some variation. The object to be inspected 2 is imaged together with a size index 18 which serves as a comparison reference for the dimensions of the object to be inspected 2. As the dimension index 18, an object, a ruler, a sticker or the like having a known dimension is used.
The image pickup is performed by installing the CCD camera 3 as the image pickup means 7 on the inspection object 2. Alternatively, a VTR taken by an ordinary video camera may be brought back and reproduced.
Alternatively, the subject 2 to be inspected is photographed, and this photograph is taken by the CCD.
The image is taken by the camera 3. Alternatively, it is transmitted and input remotely via the communication modem 10. As a result, the image memory 17 stores the grayscale image information of the inspection object 2.

The deformed portion extracting means 4 first processes the grayscale image information in the image memory 17 with grayscale morphology. By way along in figure, gray-scale image is gray information in a matrix
Since the information is lined up, the position can be moved vertically or horizontally.
Looking at the grayscale information while moving, the horizontal axis is displayed as shown in Fig. 2.
In the graph with brightness on the vertical axis and the vertical axis,
can get. When the gradation change curve shown in FIG. 2 (a) is obtained from the original image, the part having a valley change is a deformed portion candidate if the width of the valley is within a predetermined range. Be recognized. The portion indicated by the broken line in FIG. 2B is a deformed portion.
It is a candidate . Image of flattening the valleys of the deformation candidate
By doing it on the body, a filtered image is obtained. This file
For the image, the change curve of the light and shade corresponding to Fig. 2 (a)
Is as shown in FIG. FIG. 2D shows a gradation change curve corresponding to an inverted version of the original image after subtracting the filtered image. In FIG. 2D, it is obtained from the original image.
Only the valley-shaped deformed portion candidates of the gradation curve are characteristically shown. The presence or absence of a valley-shaped partially deformed portion is detected by comparing this with a predetermined threshold value. The same applies to the mountain-shaped partially deformed portion. Such processing is performed on the entire image.

Next, the deformed portion extracting means 4 performs calibration so that the dimension of the dimension index 18 in the image becomes the dimension of the actual dimension index. By the calibration, the vertical and horizontal dimensions of the detected partial deformed portion are obtained in actual size. Therefore, the deformed portion extraction means 4 is configured to detect each detected partial
The area of the deformed portion, the number of the partially deformed portions having a predetermined area or more, and the area of the entire partially deformed portion occupies the area of the entire image
The deterioration rate determining means 6 is calculated as the deformed portion data.
Send to

On the other hand, the statistic means 5 carries out a normalization pre-processing so that the gradation of light and shade in the original image shown in FIG. 3 (a) becomes a constant value. In FIG. 3A, the gradations of five levels shown in the legend of FIG. 3E are distributed, and a dark deformed portion is seen over a wide area from the lower side to the right side of the image.
There is a darkest deformed portion at the positions of coordinates J7, K1 and K2. Due to the normalization, an image with extremely high contrast has low contrast. In the normalized image shown in FIG. 3B, the lightest part in FIG. 3A has a slightly darker color. Then, as shown in FIG. 3C, the screen is divided into small regions at predetermined vertical and horizontal intervals. Each small area includes a fixed number of pixels. The statistical unit 5 calculates, for each small area, the average value and variance of the gray values of the pixels included in the small area. Also, the average value of the statistics of each of these small areas,
Calculate the variance. The statistical amount is sent to the deterioration degree determining means 6.

Deterioration degree judging means 6 determines from the deformed portion data
The degree of progress of deterioration of the object to be inspected is determined according to predetermined determination criteria and rules. The criteria and rules are as follows, for example.

(1) When there is a partially deformed portion whose area is equal to or larger than a predetermined value, it is determined that there is deterioration.

(2) When there is a predetermined number or more of partially deformed portions having an area of a predetermined value or more, it is determined that there is deterioration.

(3) When the ratio of the area of the entire partially deformed portion to the entire image exceeds a predetermined value, it is determined that there is deterioration.

Here, the judgment criteria such as the predetermined value of the area, the predetermined value of the number, the predetermined value of the ratio, etc. may be fixed in the deterioration degree judging means 6, but of various values in the judgment database 11. The judgment reference value is stored and can be read out in the deterioration degree judging means 6 as appropriate.

Next, the deterioration degree judging means 6 calculates
The degree of progress of deterioration of the object to be inspected is determined according to predetermined determination criteria and rules. More specifically, first, it is determined whether each small area is a statistically deformed portion . The criteria and rules are as follows, for example.

(1) Small areas with large variance are small shades
This is a statistically deformed part with a bias of.

(2) A small area whose variance is small but whose average value greatly differs from the average value in the entire image is a statistically deformed portion in which the unevenness of the shade covers the entire image.

(3) A small area having a small variance and an average value close to the average value of the entire image is not a statistically deformed portion .

FIG. 3D shows the result of the above judgment. Each small area hatched in the figure is shown in FIG.
As shown in the legend of (f), the average value is larger than a predetermined value,
The variance is determined to be a statistically deformed portion because the variance is greater than a predetermined value or the average value and the variance are greater than the predetermined values. This determination result is substantially coincident with the result of visual inspection of the original image shown in FIG. 3 (a) and is quantitative.

Then, the deterioration degree judging means 6 uses a statistical change.
The number of copies is counted. For example, if the number is equal to or larger than a predetermined value, it is determined that there is deterioration. Here, the judgment criteria such as the limit of the size of the variance, the limit of the difference of the average values, the limit of the number of statistically deformed parts , etc.
Although it may be fixed in the deterioration degree judging means 6, judgment reference values of various values are stored in the judgment database 11 and can be read out in the deterioration degree judging means 6 as appropriate.

Although the determination based on the deformed portion data and the determination based on the statistic have been described above, the deterioration degree determination means 6 further determines the degree of deterioration progress of the inspection object by integrating these determinations. . The determination result is displayed on the display unit 13 and is printed out on the printer 16 as necessary. At the same time, the determination result, the deformed portion data, and the statistic are sent to the history database 12 as history data,
The history database 12 accumulates history data for each inspection object 2 together with an inspection date and the like.

The deterioration degree determining means 6 is the history database 1
The history data and inspection date accumulated in 2 are read out, the change curve of the history data is approximately determined, and the future transition of the history data is predicted. Alternatively, by applying the historical data to the change curve examined from the results of the actual results and the acceleration test, the future transition of the historical data is predicted. This allows
You can know in advance when the repair coating will be required.

In the present embodiment, the deformed portion extracting means 4, the statistical means 5 and the deterioration degree judging means 6 of the coating film deterioration diagnosing system 1 are constituted by individual units and housed in the control device 8. However, the entire control device 8 may be configured by a microcomputer or a personal computer. Further, the determination can be advanced by an interactive operation using the operation input from the mouse 14 or the keyboard 15.

[0044]

The present invention exhibits the following excellent effects.

(1) Since the deterioration can be quantitatively detected, there is no individual difference in the conventional visual inspection.

(2) The deformed portion can be accurately detected without being affected by shading.

(3) Since the statistic amount is obtained by dividing into small areas, it is possible to detect even if the deformed portion spreads over a wide range.

(4) Since deterioration can be predicted, the time for repair painting can be known in advance.

[Brief description of drawings]

FIG. 1 is a block diagram of a coating film deterioration diagnosis system showing an embodiment of the present invention.

FIG. 2 is an explanatory diagram of a grayscale morphology process applied to the present invention.

FIG. 3 is an operation explanatory diagram of a statistical means and a deterioration degree determination means according to the present invention.

[Explanation of symbols]

 1 Coating Film Degradation Diagnosis System 4 Deformation Part Extraction Means 5 Statistical Means 6 Degradation Degree Judging Means 7 Imaging Means

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Sadao Degawa Inventor Sadao Izumi 3-15-15 Toyosu, Koto-ku, Tokyo Ishikawa Shimaharima Heavy Industries Co., Ltd. Toni Technical Center (72) Inventor Yukihiro Kono 3-chome, Toyosu, Koto-ku, Tokyo No. 1-15 Ishikawa Shima-Harima Heavy Industries Co., Ltd. Toni Technical Center (72) Inventor Teruzou Sugano 7-32-17 Koganehara, Matsudo-shi, Chiba (56) Reference JP 3-2649 (JP, A) Actual Kaihei 3-125245 (JP, U)

Claims (2)

(57) [Claims] 1. A split deterioration of the coating steel building steel bridges etc.
It is, peeling, in order to test based on the Deformation portion consisting of leaching steel rust, imaging the inspection object as a gray-scale image, this
Method for Diagnosis of Degradation of Coating Film by Detecting Distorted Area from Image Information
, The gray value is changed to dark or light from the gray image information above.
Even if it is turned on, the width until returning to the original is within the predetermined threshold.
The filter image is formed by averaging the gray values in the biased area.
From the image obtained by subtracting the image from the original image, above the specified gray level threshold
Rutotomoni strange-shaped portion data to detect the part as a partial strange-shaped portion
The area of each partially deformed part as a
The number of deformed parts and the partial deformed part with respect to the area of the entire image
The ratio occupied by the entire area is obtained, the grayscale image information is divided into small areas at predetermined intervals, and a statistical amount consisting of the average value and variance of the grayscale values contained in the small area is obtained for each small area. average be smaller dispersion and dispersion of larger small region is determined to statistically Henjo portion and a greatly different small regions from the average value of the entire image from the amount, the number of the statistical Henjo unit is equal to or more than a predetermined value Then, it is determined that there is deterioration, and on the other hand, from the deformed portion data, there is a partial deformed portion whose area is equal to or larger than a predetermined value, or whether there are a predetermined number of partial deformed portions whose area is equal to or larger than a predetermined value Or, the area of the entire image is partially deformed
When the ratio of the body area exceeds a predetermined value, it is determined that there is deterioration, and the degree of progress of deterioration of the inspection target is determined by combining the deterioration determination from these statistics and the deterioration determination from the deformed portion data. A method for diagnosing coating film deterioration, which comprises: 2. A split deterioration of the coating steel building steel bridges etc.
In order to inspect based on the deformed portion formed by peeling and rust leaching of the steel material, imaging means is provided for picking up an object to be inspected as grayscale image information. Detect
In the coating film deterioration diagnosing device, the gray scale is
Even if the value changes to dark or light, the width before returning to
The gray value is averaged for the biased portion with a width within the threshold.
From the image obtained by subtracting the created filtered image from the original image
Areas above a certain gray level threshold are detected as partial deformed portions, and the area of each partial deformed portion is determined as the deformed portion data.
The number of partially deformed areas that are larger than the area and the area of the entire image
On the other hand, a deformed portion extracting means for obtaining the ratio of the area of the entire partially deformed portion, and dividing the grayscale image information into small regions at a predetermined interval and averaging the grayscale values included in the small regions for each small region. Statistical means for obtaining a statistic consisting of values and variances,
Average be smaller dispersion and dispersion of larger small region from this statistic is determined to statistically Henjo portion and a greatly different small regions from the average value of the entire image, the number of the statistical Henjo portion predetermined If the value is equal to or larger than the value, it is determined that there is deterioration, and on the other hand, from the deformed portion data, there is a partial deformed portion whose area is equal to or larger than a predetermined value, or there is a predetermined number or more of partial deformed portions whose area is equal to or larger than the predetermined value. Or, the surface of the partial deformation part with respect to the area of the entire image
When the ratio of the product exceeds a predetermined value, it is determined that there is deterioration, and the degree of deterioration that determines the degree of deterioration progress of the inspection target by combining the deterioration determination from these statistics and the deterioration determination from the deformed part data An apparatus for diagnosing coating film deterioration, comprising: a determining unit.