Crack information extraction and processing system and method

Abstract

The invention provides a crack information extraction and processing system and method, and the system comprises a power supply subsystem which is used for providing electric energy for the whole system; the acquisition system is used for acquiring image data of the to-be-monitored part in real time; the data processing subsystem is used for processing the image data acquired by the acquisition system; the positioning subsystem is used for real-time positioning; and the illumination subsystem is used for providing illumination for the acquisition system. The visual image recognition monitoring system is used for crack detection, the measurement precision is high, manual participation is not needed, the data acquisition density can be automatically set, an alternating current / solar power supply mode is supported, maintenance is avoided, the system is suitable for field application, a large-capacity data storage device is adopted, remote real-time monitoring alarm can be selected, and the system is suitable for popularization and application. And meanwhile, all-day uninterrupted work in the daytime and at night can be realized.

Description

Technical Field

[0001] The invention belongs to the field of engineering repair, and specifically relates to a crack information extraction and processing system and method. Background Art

[0002] With the acceleration of industrialization, reinforced concrete structures have been widely used in industrial production due to their advantages of low cost, high strength and good plasticity. Due to its excellent physical and mechanical properties, it has become a good building material. However, due to its heterogeneity and anisotropy, its internal defects are random, which causes the cracking problem of reinforced concrete structures in engineering. Similarly, in high and steep open-air slopes, after weathering, rock and soil break and cracks appear, there is a huge danger. At this time, it is necessary to detect the cracks so that the corresponding repair methods can be used to repair the cracks later.

[0003] At present, the main crack detection methods include ground penetrating radar detection and ultrasonic detection. However, the above detection methods have certain limitations in the specific detection process. They cannot achieve high-precision detection of small cracks, and the measurement accuracy cannot be guaranteed. In addition, the specific detection process requires the on-site participation of detection personnel. For locations that are inconvenient for detection personnel to reach, the detection process is inconvenient to implement, and unmanned detection cannot be achieved around the clock. Summary of the invention

[0004] The purpose of the present invention is to provide a crack information extraction and processing system and method, which adopts a visual image recognition monitoring system for crack detection, has high measurement accuracy, does not require human intervention, the data acquisition density can be set by itself, supports AC / solar power supply, is maintenance-free, is suitable for field applications, has a large-capacity data storage device, can be optionally equipped with remote real-time monitoring and alarm, and can achieve uninterrupted work all day and night.

[0005] In order to realize the above technical features, the object of the present invention is achieved as follows: a crack information extraction and processing system, comprising: Power supply subsystem, used to provide electrical energy for the entire system; An acquisition system, used for acquiring image data of the part to be monitored in real time; A data processing subsystem, used for processing the image data collected by the acquisition system; Positioning subsystem, used for real-time positioning; The lighting subsystem is used to provide lighting for the acquisition system.

[0006] Preferably, the power supply subsystem includes batteries and solar panels, and the power supply subsystem is simultaneously connected to the collection system, the data processing subsystem, the positioning subsystem and the lighting subsystem, and provides electrical energy.

[0007] Preferably, the acquisition system includes a camera arranged at the part to be monitored, the camera is connected to an image acquisition card via a signal line and is used for image acquisition and storage of the part to be monitored; and also includes a speed sensor.

[0008] Preferably, the data processing subsystem includes a computer, which is used for slope damage image processing and identifying slope damage types; and also includes an image storage and display module for image storage and display.

[0009] Preferably, the positioning subsystem adopts a GPS positioning system; The lighting subsystem adopts a spotlight.

[0010] Another aspect of the present invention provides a method for extracting and processing crack information. The method is implemented by using the crack information extraction and processing system, and comprises the following steps: Step 1, crack image acquisition: Arrange a collection system around the cracks, use a laser night vision heavy-duty pan-tilt camera to provide uninterrupted imaging day and night, and install solar light-sensing lighting. Set up drainage ditches within a certain range of the cracks to guide water and prevent rainwater erosion from affecting crack monitoring. Step 2, crack image preprocessing: Step 2.1, crack image grayscale processing: Performing grayscale processing on the crack image captured by the camera; Step 2.2, grayscale adaptive stretching enhancement processing of crack image: Performing adaptive stretching enhancement processing on the crack image after grayscale processing; Step 2.3, crack image binarization: For the image processed in step 2.2, a specific function is used as a standard to find the binarization threshold and perform binarization processing; Step 3, feature extraction of crack images: Step 3.1, Change of crack development azimuth: Performing azimuth transformation on the crack image after binary processing; Step 3.2, azimuth direction changes; Step 4: Test the development degree of the crack image.

[0011] The specific process in step 2.1 is: First, grayscale mapping converts the acquired image into images of different gray levels, and after histogram equalization, the output image pixels are the same; Next, the grayscale nonlinear transformation of the image mainly includes logarithmic transformation and exponential transformation. The logarithmic transformation form is: , exponential transformation form: , the grayscale distribution of the transformed image matches human vision; among them, x is the image pixel row, y is the image pixel column, g ( x , y ) is the grayscale value of the output image after stretching, f ( x , y ) is the grayscale value of the original input image; a is the image translation variation coefficient; b It is a translation-telescopic coupling variable system; c is the coefficient of expansion and contraction variation; Finally, the grayscale statistical normalization of the image is performed to adjust the grayscale mean and variance of the image to given values, thereby enhancing the image contrast and facilitating the statistical analysis of crack characteristics.

[0012] The specific process in step 2.2 is: The grayscale histogram is divided into multiple intervals to represent the crack image, and then linear processing is performed to enhance the image. f ( x , y ) is the input image, and its minimum grayscale level A and maximum grayscale level B are defined as follows: A = min[ f ( x , y )], B =min[ f ( x , y )],Will A and B Linearly mapped to 0 to 255 respectively, the final image g ( x , y )for: , according to the grayscale characteristics of the collected crack image, the above formula is improved and stretched: ; Where: x is the image pixel row, y is the image pixel column, g ( x , y ) is the grayscale value of the output image after stretching, f ( x , y ) is the gray value of the original input image, x min is the minimum grayscale value of the input image, x maxis the maximum grayscale value of the input image, z max is the maximum grayscale value of the output image; The specific process of using the specific function in step 2.3 as a standard to obtain the binarization threshold is as follows: Suppose the total number of pixels of the collected crack image after grayscale processing is N , the maximum gray level of all pixels is L , then the grayscale size of the entire image is in the interval [0, L -1], the gray value is i Number of pixels n i ,but i The probability is: ; Assume a threshold value of T , the gray value is in [0, T -1] is the C0 area, and the gray value is in [ T , L -1] area is C1, then the probability of C0 appearing is: ; The probability of C1 appearing is: ; So the grayscale means of C0 and C1 are: ; ; The grayscale mean of the entire image is: ; The between-class variance is defined as: ; make T In [0, L -1] range in increments of 1, and when σ 2 At maximum, the corresponding T The value is the optimal threshold for binarization.

[0013] The specific process in step 3.1 is: The azimuth change of the slope crack curve, that is, the azimuth change of the development of the new and old crack curves, is defined as: Set point a The coordinates are ( x 1, y 1) Point b for( x 2, y 2), let the angle be ,but , , ; The value indicates the degree of deflection of the crack curve. The larger the value, the greater the deflection, and the smaller the value, the less deflection. The specific process in step 3.2 is: Due to the sliding block generated by the landslide disaster body falling down, the crack position of the new landslide crack curve is vertically upward compared with the old landslide crack curve. Therefore, the azimuth change range of the new and old crack curves is -90º~+90º, and the azimuth change is specified as negative if it deviates to the left and positive if it deviates to the right.

[0014] The specific process in step 4 is: The displacement of the crack curve reflects the sliding distance of the curve in the plane space, which is defined as: in the same plane, two crack curves at different time stages are compared and analyzed, and the calibration point is selected. The image coordinate self-recognition function is used to read the left information of the calibration monitoring point, and the point is set. a The coordinates are ( x 1, y 1) Point b for( x 2, y 2), then the displacement of crack development is: .

[0015] The present invention has the following beneficial effects: 1. Aiming at the great danger posed by weathering and cracks in rock and soil fragmentation in high and steep open-air slopes, the present invention innovatively proposes a method for extracting and processing crack information, combining the characteristics of crack images of slopes with potential landslide risks, and through the study of image denoising and image enhancement algorithms, a slope crack image preprocessing technology is developed.

[0016] 2. The present invention adopts the crack image grayscale adaptive stretching enhancement processing, performs grayscale processing according to the characteristics of the crack image, and improves and stretches the grayscale calculation formula.

[0017] 3. The crack image binarization proposed in the present invention utilizes the maximum intra-class variance Otsu binarization algorithm to digitize the split image information and store it in a computer.

[0018] 4. The crack image feature extraction proposed in the present invention includes azimuth change and azimuth change direction, and proposes measurement indicators respectively and gives representative significance.

[0019] 5. The crack image proposed in the present invention is used to test the development degree, and the crack curves at different time stages are compared and analyzed. Calibration points are selected on the crack curves, and the image coordinate self-recognition function is used to read the calibration monitoring point information to calculate the crack displacement information. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0021] Figure 1 This is a composition diagram of the slope crack monitoring system involved in the present invention.

[0022] Figure 2 This is a schematic diagram of the slope crack monitoring system of the present invention.

[0023] Figure 3 This is the image of the crack after grayscale processing.

[0024] Figure 4 This is the image after grayscale adaptive stretching enhancement processing of the crack image.

[0025] Figure 5 This is the crack image after binarization. DETAILED DESCRIPTION

[0026] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0027] Embodiment 1: like Figure 1-2 As shown, a crack information extraction and processing system includes: a power supply subsystem for providing power to the entire system; a collection system for real-time collection of image data of the monitored part; a data processing subsystem for processing the image data collected by the collection system; a positioning subsystem for real-time positioning; and a lighting subsystem for providing lighting to the collection system. By adopting the above crack information extraction and processing system, it is possible to collect images of slope cracks or concrete structure cracks, and process the collected images, thereby achieving the purpose of crack information extraction. In addition, the data collection density can be set by itself, supports AC / solar power supply, is maintenance-free, suitable for field applications, has a large-capacity data storage device, and can be equipped with remote real-time monitoring and alarm, and can also achieve uninterrupted work all day and night.

[0028] Furthermore, the power supply subsystem includes batteries and solar panels. The power supply subsystem is simultaneously connected to the data collection system, the data processing subsystem, the positioning subsystem and the lighting subsystem, and provides electrical energy.

[0029] Preferably, the acquisition system includes a camera arranged at the part to be monitored, the camera is connected to an image acquisition card via a signal line and is used for image acquisition and storage of the part to be monitored; and also includes a speed sensor.

[0030] Furthermore, the data processing subsystem includes a computer, which is used for processing slope damage images and identifying slope damage types; and also includes an image storage and display module for image storage and display.

[0031] Furthermore, the positioning subsystem adopts a GPS positioning system; the positioning subsystem can be used to perform real-time positioning of the crack information extraction and processing system, so as to facilitate subsequent real-time collection of the crack position.

[0032] Furthermore, the lighting subsystem adopts a spotlight.

[0033] Embodiment 2: See also Figure 3-5 Another aspect of the present invention provides a method for extracting and processing crack information. The method is implemented by using the crack information extraction and processing system, and comprises the following steps: Step 1, crack image acquisition: Arrange a collection system around the cracks, use a laser night vision heavy-duty pan-tilt camera to provide uninterrupted imaging day and night, and install solar light-sensing lighting. Set up drainage ditches within a certain range of the cracks to guide water and prevent rainwater erosion from affecting crack monitoring. Step 2, crack image preprocessing: Step 2.1, crack image grayscale processing: First, grayscale mapping converts the acquired image into images of different gray levels, and after histogram equalization, the output image pixels are the same; Next, the grayscale nonlinear transformation of the image mainly includes logarithmic transformation and exponential transformation. The logarithmic transformation form is: , exponential transformation form: , the grayscale distribution of the transformed image matches human vision; among them, x is the image pixel row, y is the image pixel column, g ( x , y ) is the grayscale value of the output image after stretching, f ( x , y ) is the grayscale value of the original input image; a is the image translation variation coefficient; b It is a translation-telescopic coupling variable system; c is the coefficient of expansion and contraction variation; Finally, the grayscale statistical normalization of the image is performed to adjust the grayscale mean and variance of the image to given values, thereby enhancing the image contrast and facilitating the statistical analysis of crack characteristics. Step 2.2, grayscale adaptive stretching enhancement processing of crack image: The grayscale histogram is divided into multiple intervals to represent the crack image, and then linear processing is performed to enhance the image. f ( x , y ) is the input image, and its minimum grayscale level A and maximum grayscale level B are defined as follows: A = min[ f ( x , y )], B =min[ f ( x , y )],Will A and B Linearly mapped to 0 to 255 respectively, the final image g ( x , y )for: , according to the grayscale characteristics of the collected crack image, the above formula is improved and stretched: ; Where: x is the image pixel row, y is the image pixel column, g ( x , y ) is the grayscale value of the output image after stretching, f ( x , y ) is the gray value of the original input image, x min is the minimum grayscale value of the input image, x max is the maximum grayscale value of the input image, z max is the maximum grayscale value of the output image; Step 2.3, crack image binarization: The binary method is used to enhance the detailed information of the crack part, providing an image processing algorithm for the next step of crack feature extraction. At the same time, the error accuracy of crack development length monitoring can be controlled at the millimeter level. The details are as follows: The Otsu algorithm uses a specific function as a standard to find the binarization threshold.

[0034] Suppose the total number of pixels of the collected crack image after grayscale processing is N , the maximum gray level of all pixels isL , then the grayscale size of the entire image is in the interval [0, L -1], the gray value is i Number of pixels n i ,but i The probability is: ; Assume a threshold value of T , the gray value is in [0, T -1] is the C0 area, and the gray value is in [ T , L -1] area is C1, then the probability of C0 appearing is: ; The probability of C1 appearing is: ; So the grayscale means of C0 and C1 are: ; ; The grayscale mean of the entire image is: ; The between-class variance is defined as: ; make T In [0, L -1] range in increments of 1, and when σ 2 At maximum, the corresponding T The value is the optimal threshold for binarization.

[0035] Step 3, feature extraction of crack images: Step 3.1, Change of crack development azimuth: The azimuth change of the slope crack curve, that is, the azimuth change of the development of the new and old crack curves, is defined as: Set point a The coordinates are ( x 1, y 1) Point b for( x 2, y 2), let the angle be ,but , , ; The value indicates the degree of deflection of the crack curve. The larger the value, the greater the deflection, and the smaller the value, the less deflection. Step 3.2, azimuth direction changes; Because the sliding block generated by the landslide hazard body falls down, the new landslide crack curve is vertically upward compared with the old landslide crack curve. Therefore, the azimuth change range of the new and old crack curves is -90º~+90º. It is stipulated that the azimuth change is negative when it deviates to the left and positive when it deviates to the right. Step 4: Test the development degree of the crack image.

[0036] The displacement of the crack curve reflects the sliding distance of the curve in the plane space, which is defined as: in the same plane, two crack curves at different time stages are compared and analyzed, and the calibration point is selected. The image coordinate self-recognition function is used to read the left information of the calibration monitoring point, and the point is set. a The coordinates are ( x 1, y 1) Point b for( x 2, y 2), then the displacement of crack development is: .

Claims

1. A crack information extraction and processing system, characterized in that: include: Power supply subsystem, used to provide electrical energy for the entire system; An acquisition system, used for acquiring image data of the part to be monitored in real time; A data processing subsystem, used for processing the image data collected by the acquisition system; Positioning subsystem, used for real-time positioning; The lighting subsystem is used to provide lighting for the acquisition system.

2. A crack information extraction and processing system according to claim 1, characterized in that: The power supply subsystem includes storage batteries and solar panels. The power supply subsystem is connected to the collection system, the data processing subsystem, the positioning subsystem and the lighting subsystem at the same time, and provides electrical energy.

3. A crack information extraction and processing system according to claim 1, characterized in that: The acquisition system includes a camera arranged at the part to be monitored, the camera is connected to an image acquisition card through a signal line and is used for image acquisition and storage of the part to be monitored; and also includes a speed sensor.

4. A crack information extraction and processing system according to claim 1, characterized in that: The data processing subsystem includes a computer, which is used for slope damage image processing and identifying slope damage types; and also includes an image storage and display module for image storage and display.

5. A crack information extraction and processing system according to claim 1, characterized in that: The positioning subsystem adopts the GPS positioning system; The lighting subsystem adopts a spotlight.

6. A method for extracting and processing crack information, characterized in that: The method is implemented by using the crack information extraction and processing system according to any one of claims 1 to 5, and comprises the following steps: Step 1, crack image acquisition: Arrange a collection system around the cracks, use a laser night vision heavy-duty pan-tilt camera to provide uninterrupted imaging day and night, and install solar light-sensing lighting. Set up drainage ditches within a certain range of the cracks to guide water and prevent rainwater erosion from affecting crack monitoring. Step 2, crack image preprocessing: Step 2.1, crack image grayscale processing: Performing grayscale processing on the crack image captured by the camera; Step 2.2, grayscale adaptive stretching enhancement processing of crack image: Performing adaptive stretching enhancement processing on the crack image after grayscale processing; Step 2.3, crack image binarization: For the image processed in step 2.2, a specific function is used as a standard to find the binarization threshold and perform binarization processing; Step 3, feature extraction of crack images: Step 3.1, Change of crack development azimuth: Performing azimuth transformation on the crack image after binary processing; Step 3.2, azimuth direction changes; Step 4: Test the development degree of the crack image.

7. A crack information extraction and processing method according to claim 6, characterized in that: The specific process in step 2.1 is: First, grayscale mapping converts the acquired image into images of different gray levels, and after histogram equalization, the output image pixels are the same; Next, the grayscale nonlinear transformation of the image mainly includes logarithmic transformation and exponential transformation. The logarithmic transformation form is: , exponential transformation form: ,The grayscale distribution of the transformed image matches human vision; in, x is the image pixel row, y is the image pixel column, g ( x , y ) is the grayscale value of the output image after stretching, f ( x , y ) is the grayscale value of the original input image; a is the image translation variation coefficient; b It is a translation-telescopic coupling variable system; c is the coefficient of expansion and contraction variation; Finally, the grayscale statistical normalization of the image is performed to adjust the grayscale mean and variance of the image to given values, thereby enhancing the image contrast and facilitating the statistical analysis of crack characteristics.

8. A crack information extraction and processing method according to claim 7, characterized in that: The specific process in step 2.2 is: The grayscale histogram is divided into multiple intervals to represent the crack image, and then linear processing is performed to enhance the image. f ( x , y ) is the input image, and its minimum grayscale level A and maximum grayscale level B are defined as follows: A = min[ f ( x , y )], B =min[ f ( x , y )],Will A and B Linearly mapped to 0 to 255 respectively, the final image g ( x , y )for: , according to the grayscale characteristics of the collected crack image, the above formula is improved and stretched: ； Where: x is the image pixel row, y is the image pixel column, g ( x , y ) is the grayscale value of the output image after stretching, f ( x , y ) is the gray value of the original input image, x min is the minimum grayscale value of the input image, x max is the maximum grayscale value of the input image, z max is the maximum grayscale value of the output image; The specific process of using the specific function in step 2.3 as a standard to obtain the binarization threshold is as follows: Suppose the total number of pixels of the collected crack image after grayscale processing is N , the maximum gray level of all pixels is L , then the grayscale size of the entire image is in the interval [0, L -1], the gray value is i Number of pixels n i ,but i The probability is: ； Assume a threshold value of T , the gray value is in [0, T -1] is the C0 area, and the gray value is in [ T , L -1] area is C1, then the probability of C0 appearing is: ； The probability of C1 appearing is: ； So the grayscale means of C0 and C1 are: ； ； The grayscale mean of the entire image is: ； The between-class variance is defined as: ； make T In [0, L -1] range in increments of 1, and when σ 2 At maximum, the corresponding T The value is the optimal threshold for binarization.

9. A crack information extraction and processing method according to claim 8, characterized in that: The specific process in step 3.1 is: The azimuth change of the slope crack curve, that is, the azimuth change of the development of the new and old crack curves, is defined as: Set point a The coordinates are ( x 1, y 1) Point b for( x 2, y 2), let the angle be ,but , , ; The value indicates the degree of deflection of the crack curve. The larger the value, the greater the deflection, and the smaller the value, the less deflection. The specific process in step 3.2 is: Due to the sliding block generated by the landslide disaster body falling down, the crack position of the new landslide crack curve is vertically upward compared with the old landslide crack curve. Therefore, the azimuth change range of the new and old crack curves is -90º~+90º, and the azimuth change is specified as negative if it deviates to the left and positive if it deviates to the right.

10. A crack information extraction and processing method according to claim 9, characterized in that: The specific process in step 4 is: The displacement of the crack curve reflects the sliding distance of the curve in the plane space, which is defined as: in the same plane, two crack curves at different time stages are compared and analyzed, and the calibration point is selected. The image coordinate self-recognition function is used to read the left information of the calibration monitoring point, and the point is set. a The coordinates are ( x 1, y 1) Point b for( x 2, y 2), then the displacement of crack development is: 。

Images