Anti-interference visual inspection system of chemical explosion-proof inspection robot

Abstract

The invention relates to the technical field of visual inspection, and discloses an anti-interference visual inspection system of a chemical explosion-proof inspection robot, which comprises a visual sensor, an image acquisition module, an image analysis module, an image processing module and a computer which are connected in sequence, the visual sensor, the image acquisition module, the image analysis module and the image processing module are respectively connected with the computer through wireless communication; the image acquisition module comprises image acquisition equipment and is used for acquiring images of the inspection site according to signals of the visual sensor and sending the acquired images to the image analysis module for analysis and processing; the image analysis module is used for dividing the acquired image to obtain a plurality of sub-regions and acquiring gradient amplitudes of pixel points; and meanwhile, the feature enhancement effect of the processing flow is irrelevant to the position of the reflective area in the inspection area, so that the adaptability to the influence of complex and changing ambient light is improved, and the effect of self-adaptive light interference resistance is achieved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of visual detection, in particular to an anti-interference visual detection system of a chemical explosion-proof inspection robot. BACKGROUND

[0003] The existing visual detection system of a chemical explosion-proof inspection robot has the following problems: the oil and chemical plant area is complex, and the influence of the changing environmental light on the visual detection is great, there is no effective adaptive anti-light interference method at present, and due to the different scales of the detail information in the image obtained by the camera, the image cannot be effectively divided and locally histogram equalized by using fixed-size parameters, the effect of the local histogram equalization of the image is not ideal, and the robot inspection effect is not ideal. SUMMARY

[0004] In order to solve the problems of the existing visual detection system of a chemical explosion-proof inspection robot, the oil and chemical plant area is complex, and the influence of the changing environmental light on the visual detection is great, there is no effective adaptive anti-light interference method at present, and due to the different scales of the detail information in the image obtained by the camera, the image cannot be effectively divided and locally histogram equalized by using fixed-size parameters, the effect of the local histogram equalization of the image is not ideal, and the robot inspection effect is not ideal, the application provides an anti-interference visual detection system of a chemical explosion-proof inspection robot.

[0005] The anti-interference visual detection system of a chemical explosion-proof inspection robot provided by the application adopts the following technical scheme:

[0006] The anti-interference visual detection system of a chemical explosion-proof inspection robot comprises a visual sensor, an image acquisition module, an image analysis module, an image processing module and a computer connected in sequence, and the visual sensor, the image acquisition module, the image analysis module and the image processing module are connected with the computer through wireless communication.

[0007] The image acquisition module comprises an image acquisition device, which is used for acquiring the inspection site image according to the signal of the visual sensor and sending the acquired image to the image analysis module for analysis and processing.

[0008] The image analysis module is used for dividing the acquired image to obtain a plurality of sub-regions, acquiring the gradient amplitude of the pixel points, and obtaining the gray uniformity and the edge pixel points of the sub-regions according to the gray value difference and the gradient amplitude difference of the pixel points in the sub-regions.

[0009] The image processing module is configured to effectively enhance the reflection area of the inspection site, make the contour edge clearer, make the features of the reflection area more obvious, make the reflection area easier to distinguish and identify, and better adapt to the influence of complex and changing ambient light, so as to obtain a processed image.

[0010] The computer is configured to control the robot to complete the chemical explosion-proof inspection work according to the final processed image.

[0011] Preferably, the image acquisition module comprises an image acquisition device and a light source, and the light source is configured to provide illumination to enable the image acquisition device to obtain clearer images.

[0012] Preferably, the image acquisition device comprises at least a camera configured to acquire images, the camera is arranged at a height of 1 / 2 to 2 / 3 of the inspection robot from the bottom, the optical axis of the camera passes through the longitudinal axis of the single crystal silicon rod, and the elevation angle of the camera is 15° to 45°.

[0013] Preferably, the image analysis module comprises a gray level acquisition unit and a pixel point acquisition unit, the gray level acquisition unit is configured to divide the acquired image into a plurality of sub-regions, and the pixel point acquisition unit is configured to acquire the gradient amplitude of the pixel points, and obtain the gray level uniformity of the sub-region according to the difference in the gray level and the difference in the gradient amplitude of the pixel points in the sub-region.

[0014] Preferably, the gray level acquisition unit specifically acquires the gray level in the following manner: the image of the inspection site is uniformly divided into B regions of the same size, denoted as sub-regions of the image of the inspection site, wherein B is a preset quantity parameter.

[0015] Preferably, the pixel point acquisition unit specifically acquires the gradient amplitude of the pixel points in the following manner: the Sobel operator is used to acquire the gradient amplitude of each pixel point in the image of the inspection site; a first accumulation value and a second accumulation value are respectively obtained according to the difference in the gray level and the difference in the gradient amplitude of the pixel points in the sub-region; and the gray level uniformity of the sub-region is obtained according to the first accumulation value and the second accumulation value, and the first accumulation value and the second accumulation value are inversely proportional to the gray level uniformity of the sub-region.

[0016] Preferably, the image processing module comprises an image threshold processing unit, an image erosion processing unit and an image dilation processing unit, the image threshold processing unit is configured to perform binary image threshold processing on the image to obtain a foreground mask, the image erosion processing unit is configured to perform one or more times of erosion processing on the foreground mask to obtain an erosion image, and the image dilation processing unit is configured to perform one or more times of dilation processing on the erosion image to obtain a dilation image.

[0017] Preferably, the foreground mask is subjected to one or more erosion treatments, specifically: the value after the erosion treatment is the dimension of the convolution kernel, a convolution kernel is formed, and a convolution operation is performed on the foreground mask pixel by pixel, the anchor point of the convolution kernel is scanned for each pixel of the foreground mask, and the minimum value of the foreground mask in the convolution kernel region is used to replace the anchor point pixel, thereby realizing the erosion of the image.

[0018] Preferably, the eroded image is subjected to one or more dilation treatments, specifically: the value after the dilation treatment, a convolution kernel is formed, and a convolution operation is performed on the eroded image pixel by pixel, the anchor point of the convolution kernel is scanned for each pixel of the eroded image, and the maximum value of the eroded image in the convolution kernel region is used to replace the anchor point pixel, thereby realizing the dilation of the image.

[0019] Compared with the prior art, the present application has the following advantages:

[0020] The anti-interference visual detection system of the chemical explosion-proof inspection robot provided by the present application can effectively retain the detail information of the image under different scales, so that the inspection robot can effectively use the enhanced image in the inspection area for visual detection, thereby improving the inspection effect; the processing procedure including the binary image threshold processing, the erosion processing and the dilation processing of the image processing module can effectively enhance the reflective region in the inspection, and the contour edge becomes clearer, so that the characteristics of the reflective region are more obvious and easier to distinguish and identify; meanwhile, the feature enhancement effect of the processing procedure is independent of the position of the reflective region in the inspection area, thereby improving the adaptability to the influence of the complex and changing environmental light, and achieving the effect of self-adaptive anti-light interference. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 FIG. 1 is a structural schematic diagram of an anti-interference visual detection system of a chemical explosion-proof inspection robot according to the present application.

[0022] Legend: 1, visual sensor; 2, image acquisition module; 3, image analysis module; 4, image processing module; 5, computer; 6, image acquisition device; 7, light source; 8, camera; 9, gray scale acquisition unit; 10, pixel point acquisition unit; 11, image threshold processing unit; 12, image erosion processing unit; 13, image dilation processing unit. DETAILED DESCRIPTION

[0023] Clearly, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0024] Please refer to Figure 1 The present application provides the following embodiments:

[0025] An anti-interference visual detection system of a chemical explosion-proof inspection robot, comprising a visual sensor 1, an image acquisition module 2, an image analysis module 3, an image processing module 4 and a computer 5 connected in sequence, the visual sensor 1, the image acquisition module 2, the image analysis module 3 and the image processing module 4 are connected with the computer 5 through wireless communication.

[0026] The image acquisition module 2 is used for collecting the inspection site image according to the signal of the visual sensor 1, and sending the collected image to the image analysis module 3 for analysis and processing.

[0027] The image analysis module 3 is used for dividing the collected image to obtain a plurality of sub-regions, obtaining the gradient amplitude of the pixel points, and obtaining the gray uniformity and the edge pixel points of the sub-regions according to the gray value difference and the gradient amplitude difference of the pixel points in the sub-regions.

[0028] The image processing module 4 is used for effectively enhancing the reflection area of the inspection site, making the contour edge clearer, so that the characteristics of the reflection area are more obvious, and it is easier to distinguish and identify, and it can better adapt to the influence of complex and changing environmental light, and finally obtain the processed image.

[0029] The computer is used for controlling the robot to complete the chemical explosion-proof inspection work according to the finally processed image.

[0030] The image acquisition module 2 comprises an image acquisition device 6 and a light source 7, and the light source 7 is used for providing illumination to make the image acquisition device obtain clearer images.

[0031] The image acquisition device 6 at least comprises a camera 8 used for collecting images, and the camera 8 is arranged at 1 / 2-2 / 3 height from the bottom of the inspection robot, and the elevation angle of the camera 8 is 15°-45°.

[0032] The image analysis module 3 comprises a gray acquisition unit 9 and a pixel point acquisition unit 10, the gray acquisition unit 9 is used for dividing the collected image to obtain a plurality of sub-regions, and the pixel point acquisition unit 10 is used for obtaining the gradient amplitude of the pixel points, and obtaining the gray uniformity of the sub-regions according to the gray value difference and the gradient amplitude difference of the pixel points in the sub-regions.

[0033] The gray scale acquisition unit 9 specifically acquires in the following manner: the inspection site image is evenly divided into B regions of the same size, denoted as sub-regions of the inspection site image, wherein B is a preset quantity parameter.

[0034] The pixel point acquisition unit 10 specifically acquires in the following manner: the Sobel operator is used to acquire the gradient amplitude of each pixel point in the inspection site image; the first cumulative value and the second cumulative value are respectively obtained according to the gray value difference and the gradient amplitude difference of the pixel points in the sub-region; the gray uniformity of the sub-region is obtained according to the first cumulative value and the second cumulative value, and the first cumulative value and the second cumulative value are inversely proportional to the gray uniformity of the sub-region.

[0035] The image processing module 4 includes an image threshold processing unit 11, an image erosion processing unit 12, and an image dilation processing unit 13. The image threshold processing unit 11 is configured to perform binary image threshold processing on the image to obtain a foreground mask. The image erosion processing unit 12 is configured to perform one or more erosion processes on the foreground mask to obtain an erosion image. The image dilation processing unit 13 is configured to perform one or more dilation processes on the erosion image to obtain a dilation image.

[0036] The one or more erosion processes performed on the foreground mask specifically include: the value after the erosion process is the dimension of the convolution kernel, which forms a convolution kernel. The convolution kernel is used to perform a convolution operation on each pixel of the foreground mask, and the anchor point of the convolution kernel scans each pixel of the foreground mask. The minimum value of the foreground mask in the region of the convolution kernel is used to replace the anchor point pixel, thereby realizing the erosion of the image.

[0037] The one or more dilation processes performed on the erosion image specifically include: the value after the dilation process is the dimension of the convolution kernel, which forms a convolution kernel. The convolution kernel is used to perform a convolution operation on each pixel of the erosion image, and the anchor point of the convolution kernel scans each pixel of the erosion image. The maximum value of the erosion image in the region of the convolution kernel is used to replace the anchor point pixel, thereby realizing the dilation of the image.

[0038] The anti-interference visual detection system of the chemical explosion-proof inspection robot of the application, by setting the image acquisition module 2, the image analysis module 3 and the image processing module 4, realizes the collection, analysis and processing of the image in the inspection area, thereby improving the processing quality of the image, wherein the gray scale information of different regions in the image in the inspection site is obtained through the image analysis module 3, and the performance of the sub-regions is adjusted by using the difference in the gray scale uniformity between the sub-regions, thereby improving the accuracy of the performance of the sub-regions in describing the difference in the gray scale information of the sub-regions relative to other sub-regions, further improving the effect of merging and local enhancement of the sub-regions by using the performance of the sub-regions, effectively retaining the detail information of the image under different scales, so that the inspection robot can effectively use the enhanced image in the inspection area for visual detection, thereby improving the inspection effect; through the image processing module 4, the processing procedure including the threshold processing of the binary image, the erosion processing and the inflation processing can effectively enhance the reflection area in the inspection, so that the contour edge becomes clearer, thereby making the features of the reflection area more obvious and easier to distinguish and identify; meanwhile, the feature enhancement effect of the processing procedure is irrelevant to the position of the reflection area in the inspection area, thereby improving the adaptability to the influence of the complex and changing environmental light, and achieving the effect of self-adaptive anti-light interference.

[0039] Finally, it should be pointed out that: first, in the description of the present application, it should be pointed out that, unless otherwise specified and limited, the terms "mounting", "connection", "connection" should be understood broadly, which can be mechanical connection or electrical connection, or the communication between two elements, or direct connection, "up", "down", "left", "right" and the like are only used to represent the relative positional relationship, when the absolute position of the described object changes, the relative positional relationship may change;

[0040] Secondly: the present application discloses the structure involved in the embodiment of the present application, and other structures can refer to the usual design, and the same embodiment and different embodiments of the present application can be combined with each other under the condition of no conflict;

[0041] Finally: the above only describes the preferred embodiments of the present application, and is not used to limit the present application, and any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application should be included in the protection scope of the present application.

[0042] The above are the preferred embodiments of the present application, and are not used to limit the protection scope of the present application, therefore: any equivalent change made according to the structure, shape, principle of the present application should be included in the protection scope of the present application.

Claims

1. An anti-interference visual inspection system for a chemical explosion-proof inspection robot, characterized in that: The system includes a visual sensor (1), an image acquisition module (2), an image analysis module (3), an image processing module (4), and a computer (5) connected in sequence. The visual sensor (1), the image acquisition module (2), the image analysis module (3), and the image processing module (4) are connected to the computer (5) via wireless communication. The image acquisition module (2) is used to acquire images of the inspection site based on the signal from the vision sensor (1) and send the acquired images to the image analysis module (3) for analysis and processing. The image analysis module (3) is used to divide the acquired image into several sub-regions, obtain the gradient magnitude of the pixels, and obtain the gray uniformity and edge pixels of the sub-regions based on the difference in gray values ​​and gradient magnitude of the pixels in the sub-regions. The image processing module (4) is used to effectively enhance the reflective area of ​​the inspection site, making the outline edge clearer, thereby making the features of the reflective area more obvious, easier to distinguish and identify, and better able to adapt to the influence of complex and changing ambient light, and finally obtain the processed image. The computer is used to control the robot to complete the chemical explosion-proof inspection work based on the final processed image.

2. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 1, characterized in that: The image acquisition module (2) includes an image acquisition device (6) and a light source (7), the light source (7) being used to provide illumination so that the image acquisition device can obtain a clearer image.

3. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 2, characterized in that: The image acquisition device (6) includes at least a camera (8) for acquiring images. The camera (8) is positioned at 1 / 2 to 2 / 3 of the height of the inspection robot from the bottom, and the elevation angle of the camera (8) is 15° to 45°.

4. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 1, characterized in that: The image analysis module (3) includes a grayscale acquisition unit (9) and a pixel acquisition unit (10). The grayscale acquisition unit (9) is used to divide the acquired image into several sub-regions. The pixel acquisition unit (10) is used to acquire the gradient magnitude of the pixel and obtain the grayscale uniformity of the sub-region based on the difference in grayscale value and gradient magnitude of the pixel in the sub-region.

5. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 4, characterized in that: The specific acquisition method of the grayscale acquisition unit (9) is as follows: the inspection site image is evenly divided into B regions of the same size, which are denoted as sub-regions of the inspection site image, where B is a preset quantity parameter.

6. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 4, characterized in that: The specific acquisition method of the pixel acquisition unit (10) is as follows: the gradient magnitude of each pixel in the inspection site image is obtained by using the Sobel operator; the first accumulated value and the second accumulated value are obtained according to the difference in gray value and gradient magnitude of the pixels in the sub-region; the gray uniformity of the sub-region is obtained according to the first accumulated value and the second accumulated value, and the first accumulated value and the second accumulated value are both inversely proportional to the gray uniformity of the sub-region.

7. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 1, characterized in that: The image processing module (4) includes an image thresholding unit (11), an image erosion processing unit (12), and an image dilation processing unit (13). The image thresholding unit (11) is used to perform binarization image thresholding processing on the image to obtain a foreground mask. The image erosion processing unit (12) is used to perform one or more erosion processing on the foreground mask to obtain an eroded image. The image dilation processing unit (13) is used to perform one or more dilation processing on the eroded image to obtain a dilated image.

8. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 7, characterized in that: The foreground mask undergoes one or more erosion processes specifically as follows: the eroded value is the convolution kernel dimension, forming a convolution kernel. A convolution operation is performed on the foreground mask pixel by pixel, serving as the convolution kernel anchor point. The anchor point scans each pixel of the foreground mask, and the minimum value of the foreground mask in the convolution kernel region is used to replace the anchor point pixel, thereby achieving image erosion.

9. The anti-interference visual inspection system for a chemical explosion-proof inspection robot according to claim 7, characterized in that: The process of performing one or more dilation operations on the eroded image specifically includes: using the dilated values ​​to form a convolution kernel; performing a convolution operation on the eroded image pixel by pixel, with the kernel anchor point as the anchor point; scanning each pixel of the eroded image at the anchor point; and replacing the anchor point pixel with the maximum value of the eroded image in the convolution kernel region, thereby achieving image dilation.

Images