CN-122023774-A - Back splint wearing condition detection method, device, equipment and storage medium

CN122023774ACN 122023774 ACN122023774 ACN 122023774ACN-122023774-A

Abstract

The application provides a detection method, a device, equipment and a storage medium for the wearing condition of a back clip, wherein the detection method is based on the fact that an image detection target in a mine scene is a small-size target object related to the back clip (such as a small back strap when the back clip is worn on the front surface, a small part area of the back clip can only be exposed in an image due to shielding caused by changeable personnel posture, and the like), and an image enhancement mechanism which is favorable for highlighting the small-size target object in an image to be detected is specifically designed, so that the target detection accuracy of the image target of the back clip can be effectively improved through target detection of a target enhanced image after image enhancement.

Inventors

LV XIAO
LI YUANYUAN
LI RUOXI
ZHANG YUANSHENG
Chen Suohan
Qiao Haiqing
Song Yuemen
LIU PENG
ZHU MING
GUO QINGJIE

Assignees

北京北矿智能科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260202

Claims (10)

1. A method for detecting the wearing condition of a back clip, the method comprising: Collecting an image to be detected of an operator through an image collecting device arranged in a downhole operation area; sequentially executing multi-level image enhancement processing on the image to be detected according to multi-level image enhancement processing adapted by a back clamp to obtain a target enhanced image, wherein the multi-level image enhancement processing comprises image quality enhancement processing, illumination and color correction processing, defogging recovery processing based on dark channel priori and edge enhancement processing; And inputting the target enhanced image into a target detection model, outputting a back splint detection result of the target enhanced image through the target detection model, and determining the back splint wearing condition of the operator according to the back splint detection result.
2. The detection method according to claim 1, wherein the performing the multi-level image enhancement processing on the image to be detected sequentially according to the multi-level image enhancement processing adapted by the back clip, to obtain a target enhanced image, includes: performing image quality enhancement processing on the image to be detected to obtain an image to be detected with enhanced image quality as a first enhanced image; Performing illumination and color correction processing on the first enhanced image to obtain an illumination and color corrected first enhanced image serving as a second enhanced image; Performing defogging recovery processing based on dark channel prior on the second enhanced image to obtain a defogged recovered second enhanced image as a third enhanced image; And carrying out edge enhancement processing on the third enhanced image to obtain an edge enhanced third enhanced image serving as the target enhanced image.
3. The method according to claim 2, wherein the performing image quality enhancement processing on the image to be detected to obtain the image to be detected with enhanced image quality as the first enhanced image includes: dynamically adjusting target parameters corresponding to different brightness areas respectively according to brightness distribution conditions in the image to be detected, and performing self-adaptive gamma correction processing on the image to be detected according to the target parameters to obtain an initial enhanced image; and denoising the initial enhanced image in a bilateral filtering mode to obtain the first enhanced image.
4. The method according to claim 2, wherein the performing illumination and color correction processing on the first enhanced image to obtain the first enhanced image after illumination and color correction as the second enhanced image includes: Extracting a plurality of illumination component features corresponding to the different scales from the first enhanced image through Gaussian filtering of the different scales, and carrying out weighted fusion on the plurality of illumination component features according to the weight coefficients corresponding to the different scales to obtain target illumination component features; and carrying out illumination and color correction processing on the first enhanced image according to the target illumination component characteristics and a color recovery factor to obtain the second enhanced image, wherein the color recovery factor represents the color perception weight of each pixel point in the first enhanced image.
5. The method according to claim 2, wherein said performing defogging recovery processing based on dark channel prior on the second enhanced image, to obtain a defogged recovered second enhanced image as a third enhanced image, comprises: For each pixel point in the second enhanced image, determining a color component with the smallest value as a dark channel value of the pixel point from color components respectively corresponding to the pixel points in a preset window range according to the preset window range taking the pixel point as the center; determining a dark channel image corresponding to the second enhanced image according to the dark channel value of each pixel point; According to the dark channel image and the second enhanced image, determining an atmospheric light parameter corresponding to the second enhanced image; according to the atmospheric light parameters and the second enhanced image, calculating the transmissivity corresponding to each pixel point in the second enhanced image to obtain a transmissivity image corresponding to the second enhanced image; And carrying out defogging recovery processing on the second enhanced image according to the transmissivity image and the atmospheric light parameter to obtain a defogged recovered second enhanced image serving as the third enhanced image.
6. The method according to claim 2, wherein performing edge enhancement processing on the third enhanced image to obtain an edge-enhanced third enhanced image as the target enhanced image, comprises: And carrying out Laplacian sharpening on the third enhanced image to obtain a third enhanced image subjected to Laplacian sharpening processing as the target enhanced image.
7. The method according to claim 1, wherein the determining, according to the back clip detection result, a wearing condition of the back clip of the worker includes: when the back clips do not exist in the back clip detection result, judging whether back clip wearing data matched with the face of the operator exist in a personnel back clip wearing record acquired in advance before operation in a face matching mode; If the back clamp wearing data matched with the face of the operator exists, the back clamp wearing condition of the operator is determined to be normal; And if the back clip wearing data matched with the face of the operator does not exist, determining that the back clip wearing condition of the operator is abnormal.
8. A device for detecting the wearing condition of a back splint, the device comprising: the image acquisition module is used for acquiring images to be detected of operators through an image acquisition device installed in the underground operation area; The image enhancement module is used for sequentially executing multi-level image enhancement processing on the image to be detected according to multi-level image enhancement processing adapted by the back clamp to obtain a target enhanced image, wherein the multi-level image enhancement processing comprises image quality enhancement processing, illumination and color correction processing, defogging recovery processing based on dark channel prior and edge enhancement processing; The target detection module is used for inputting the target enhanced image into a target detection model, outputting a back clamp detection result of the target enhanced image through the target detection model, and determining the back clamp wearing condition of the operator according to the back clamp detection result.
9. An electronic device comprising a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory communicating via the bus when the electronic device is in operation, the machine-readable instructions when executed by the processor performing the steps of the method for detecting a wear condition of a back splint according to any of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when executed by a processor, performs the steps of the method for detecting a wearing condition of a back clip as claimed in any one of claims 1 to 7.

Description

Back splint wearing condition detection method, device, equipment and storage medium Technical Field The application relates to the technical field of target detection, in particular to a method, a device, equipment and a storage medium for detecting wearing conditions of back clamps. Background In order to ensure the safety of operators, when the operators enter the working area under the mine, the anti-smashing safety back clip with the reflecting plate on the back is required to be worn. Wherein, in order to detect whether the operators in the underground operation area wear the anti-smashing safety back clamp, the prior art utilizes an anti-explosion camera integrated with an AI (ARTIFICIAL INTELLIGENCE ) chip and edge computing equipment, and the highlight area of the reflector is positioned in real time in the video stream collected in the underground operation area through a deep learning target detection algorithm, so as to determine whether the operators in the video stream wear the anti-smashing safety back clamp. However, because the shooting environment in the underground operation area is worse, continuous dim light, local overexposure caused by the miner lamp, high-concentration dust, water vapor and other environmental factors can all lead to the fact that the actually shot video image is more fuzzy, so that the extraction of the characteristics of the reflecting plate becomes extremely difficult, and the front surface of the anti-smashing safety back clamp is only narrower and thinner braces and does not have any reflecting elements, so that the technical defect of poor recognition precision exists in the mode of recognizing the anti-smashing safety back clamp in the image by only depending on the reflecting characteristics in the prior art is caused. Disclosure of Invention In view of this, the present application provides a method, apparatus, device and storage medium for detecting wearing conditions of a back clip, where, based on an image detection target in a mine scene, the image detection target is a small-size target object related to the back clip (for example, a small area where only the back clip can be exposed in an image due to shielding caused by a narrow back strap and a changeable posture of a person when the back clip is worn on the front surface), and an image enhancement mechanism that is favorable for highlighting the small-size target object in an image to be detected is specifically designed, so that by performing target detection on a target enhanced image after image enhancement, the accuracy of image target detection on the back clip can be effectively improved. In order to make the above objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below. In a first aspect, an embodiment of the present application provides a method for detecting a wearing condition of a back clip, where the method includes: Collecting an image to be detected of an operator through an image collecting device arranged in a downhole operation area; sequentially executing multi-level image enhancement processing on the image to be detected according to multi-level image enhancement processing adapted by a back clamp to obtain a target enhanced image, wherein the multi-level image enhancement processing comprises image quality enhancement processing, illumination and color correction processing, defogging recovery processing based on dark channel priori and edge enhancement processing; And inputting the target enhanced image into a target detection model, outputting a back splint detection result of the target enhanced image through the target detection model, and determining the back splint wearing condition of the operator according to the back splint detection result. In a second aspect, an embodiment of the present application provides a device for detecting a wearing condition of a back clip, where the device includes: the image acquisition module is used for acquiring images to be detected of operators through an image acquisition device installed in the underground operation area; The image enhancement module is used for sequentially executing multi-level image enhancement processing on the image to be detected according to multi-level image enhancement processing adapted by the back clamp to obtain a target enhanced image, wherein the multi-level image enhancement processing comprises image quality enhancement processing, illumination and color correction processing, defogging recovery processing based on dark channel prior and edge enhancement processing; the target detection module is used for inputting the target enhanced image into a target detection model, outputting a back clamp detection result of the target enhanced image through the target detection model, and determining the back clamp wearing condition of the operator according to the back clamp detection result In a third aspect,