CN117333825B - Cable bridge monitoring method based on computer vision - Google Patents

Abstract

The invention relates to the field of image processing, in particular to a cable bridge monitoring method based on computer vision, which comprises the steps of obtaining the gray level abnormality degree of a hyperspectral image channel image on the surface of each cable galvanized bridge, determining hyperspectral images with zinc plating unevenness according to the gray level abnormality degree, obtaining the abnormality degree of each pixel point in each channel image, calculating the probability that the pixel point is a zinc plating abnormality pixel point, obtaining zinc plating abnormality pixel points according to the probability, carrying out mean shift clustering on the zinc plating abnormality pixel points in each channel image, obtaining the optimal reference weight of each channel image according to each clustering result and the gray level abnormality degree, and carrying out gray level and connected domain analysis on the hyperspectral images according to the optimal reference weight to obtain the zinc plating uneven area on the surface of the cable galvanized bridge. According to the invention, the zinc plating uneven area on the surface of the cable zinc plating bridge is accurately and reliably monitored by extracting the characteristics in the image.

Description

A cable tray monitoring method based on computer vision

Technical field

This application relates to the field of computer vision, and specifically to a cable tray monitoring method based on computer vision.

Background technique

Cable tray is a product derived to protect cables. It is a protective shell for cables to prevent cables from being damaged by external factors. It is mainly composed of brackets, brackets and installation accessories. All parts of the cable tray need to be galvanized. Galvanizing can not only make the bridge more beautiful, but also play an anti-rust role. It has a good protective effect on materials that are often exposed outdoors and can improve the durability of the cable tray. Lifespan and durability. However, many factors such as the metal composition of the bridge, surface roughness, geometric dimensions of the workpiece, and hot-dip galvanizing process will affect the thickness of the galvanized layer. If the thickness of the galvanized layer is uneven, the corrosion resistance of the bridge will be reduced, and The surface of the bridge is prone to defects. The more uniform the thickness of the galvanized layer is on the bridge, the better the quality of the bridge. Therefore, it is necessary to monitor the uniformity of the galvanized layer of the cable tray to ensure the production quality of the cable tray.

Since the color of the metal on the bridge surface changes little in the images collected by traditional cameras before and after galvanizing, it is difficult to distinguish the subtle differences at different locations using ordinary cameras, which makes it difficult to evaluate the thickness of the galvanized layer at different locations.

Therefore, according to the different characteristic spectra of different metals, and the spectral intensity is also related to the content of the metal, this technical solution uses a hyperspectral camera to collect the surface image of the cable tray, and by analyzing the spectral information in the image, the zinc plating of the bridge is obtained The slight differences in various positions in the image can lead to uneven areas of galvanizing, thereby enabling monitoring of the galvanizing quality of the bridge surface.

Contents of the invention

The present invention provides a computer vision-based cable bridge monitoring method to solve the problem of insufficiently accurate monitoring of the galvanizing quality of the cable bridge surface, and adopts the following technical solution:

The present invention proposes a cable bridge monitoring method based on computer vision, which includes:

Collect the hyperspectral image of the surface of the galvanized cable bridge and perform dimensionality reduction processing to obtain the dimensionally reduced hyperspectral image;

The degree of grayscale anomaly of each channel image is obtained based on the difference between the grayscale value of each pixel in each channel image of the dimensionally reduced hyperspectral image relative to the grayscale value in the channel image;

Hyperspectral images with uneven galvanizing are determined based on the grayscale abnormality of each channel image;

According to the gray value of each pixel in each channel image in the hyperspectral image with uneven galvanizing, the abnormality degree of each pixel in each channel image is obtained;

Calculate the probability that the pixel is an abnormal zinc plating pixel based on the abnormality degree of each pixel in each channel image in the hyperspectral image with uneven zinc plating;

Determine abnormal galvanizing pixels in hyperspectral images with uneven galvanizing based on probability and probability thresholds;

Perform mean shift clustering on abnormal galvanizing pixels in each channel image of a hyperspectral image with uneven galvanizing;

According to the number of pixels in each clustering result and the abnormality degree of galvanized abnormal pixels in each clustering result in the channel image and the grayscale abnormality degree of the channel image, the maximum value of each channel image is obtained. Excellent reference weight;

According to the optimal reference weight of each channel image, the hyperspectral image is grayscale processed and connected domain analysis is performed to obtain the galvanized uneven area on the surface of the cable galvanized bridge.

Further, the dimensionality reduction processing method is as follows:

Obtain the normalized grayscale histogram of each channel image of the hyperspectral image;

The maximum grayscale value and minimum grayscale value in the grayscale histogram corresponding to each channel image and the proportion of each grayscale value in the grayscale range are used as a dimension of the vector, and the grayscale of each channel image is The histogram is converted into a vector with the same dimension as the number of gray values in the gray histogram;

Divide the channel images with the same maximum gray value and minimum gray value in each channel image into a group, and calculate the cosine similarity between the vectors corresponding to each channel image in the same group;

The channel images corresponding to the two vectors with cosine similarity values greater than or equal to the threshold are mutually redundant channel images, retain the channel image with the smallest dimension among the images that are redundant channels, and eliminate other channel images;

All images with mutually redundant channels are processed in sequence to achieve dimensionality reduction of hyperspectral images.

Further, the method for determining the presence of uneven galvanizing hyperspectral images is:

Obtain the grayscale abnormality degree of each channel image and obtain the average grayscale abnormality degree. As a hyperspectral image, there is the possibility of uneven galvanizing;

The possibility is compared with the possibility threshold. When the possibility is greater than the threshold, the hyperspectral image is a hyperspectral image with uneven galvanizing.

Further, the acquisition method of galvanized abnormal pixels in the hyperspectral image is as follows:

Obtain the abnormality degree of each pixel in the hyperspectral image in each channel image:

In the formula, For the first/> The grayscale value with the largest proportion of the channel image,/> For the first/> pixel at/> Grayscale value in channel image,/> For the first/> pixel at/> The degree of abnormality in each channel image;

Calculate the probability that each pixel in the hyperspectral image is a galvanized abnormal pixel. The calculation formula is:

In the formula, is the number of channel images of the hyperspectral image after dimensionality reduction,/> For the first/> The probability that a pixel is an abnormal galvanized pixel;

when When it is less than the probability threshold, the pixels are noise points, otherwise, the /> The pixels are galvanized abnormal pixels in the hyperspectral image.

Further, the method for obtaining the optimal reference weight of each channel image is as follows:

The grayscale abnormality degree of each channel image is used as the initial reference weight of the channel image;

Mean shift clustering is performed on the coordinates of galvanized outliers in each channel image, where the The number of mean shift clustering results obtained in channel images is/> , each clustering result corresponds to an uneven area;

Then the optimal reference weight calculation formula for each channel image is:

In the formula, For the first/> The optimal reference weight of channel images,/> For the first/> Initial reference weights of channel images, /> For the first/> Channel image/> The number of pixels contained in a clustering result, that is, the area of the uneven area in the clustering result,/> is the total number of clustering results,/> For the first/> In the channel image, the The /> contained in the clustering results The galvanized abnormal pixels are at/> The degree of anomaly within each channel image.

Further, the method for obtaining the uneven galvanized area on the surface of the galvanized cable tray is:

The optimal reference weight of each channel image is used to accumulate and sum the grayscale values of the pixels in the channel image, and the hyperspectral image is grayscaled to obtain the grayscale image of the hyperspectral image;

The Seed-Filling algorithm is used to perform connected domain analysis on the grayscale image of the hyperspectral image, and the obtained connected domain is the galvanized uneven area on the surface of the cable galvanized bridge.

Further, the method for obtaining the abnormality degree includes:

According to the difference between the gray value of each pixel in each channel image in the hyperspectral image with uneven galvanizing and the gray value with the largest proportion in the channel image, the value of each pixel in each channel image is obtained. degree of abnormality.

Further, the method for obtaining the degree of grayscale anomaly includes:

The degree of grayscale anomaly of each channel image is obtained based on the variance of the grayscale value of each pixel in each channel image of the dimensionally reduced hyperspectral image relative to the average grayscale value in the channel image.

Further, the possibility threshold is set to 0.1.

Further, the probability threshold is set to 0.6.

The beneficial effects of the present invention are: using computer vision, based on hyperspectral images, by performing dimensionality reduction processing on the hyperspectral images of the surface of the galvanized cable bridge, the grayscale abnormality degree of each channel image is obtained, and the grayscale abnormality degree is determined according to the grayscale abnormality degree. To obtain a hyperspectral image with uneven galvanizing, the difference between the gray value of each pixel in each channel image and the gray value with the largest proportion in the channel image is used to obtain the value of each pixel in each channel. The degree of abnormality in the image is calculated based on the degree of abnormality. The probability that the pixel is an abnormal pixel of galvanizing is calculated. Based on the probability and threshold, the abnormal pixel of galvanizing in the hyperspectral image with uneven galvanizing is determined. For those with galvanized The galvanized abnormal pixels in each channel image of the uneven hyperspectral image are subjected to mean shift clustering. According to the number of pixels in each clustering result and the galvanized abnormal pixels in each clustering result, The degree of abnormality in the channel image and the degree of grayscale abnormality in the channel image are obtained to obtain the optimal reference weight of each channel image. Based on the optimal reference weight of each channel image, the hyperspectral image is grayscale processed and connected domain analysis is performed. , to obtain the uneven galvanized area on the surface of the galvanized cable bridge, and the monitoring method is accurate and reliable.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without exerting any creative effort.

Figure 1 is a schematic flow chart of a cable tray monitoring method based on computer vision of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

An embodiment of a cable tray monitoring method based on computer vision of the present invention, as shown in Figure 1, includes:

Step 1: Collect hyperspectral images of the surface of the galvanized cable tray.

Step 2: Dimensionality reduction processing to obtain the dimensionally reduced hyperspectral image.

A hyperspectral camera is used to collect surface hyperspectral images of galvanized bridges, and the hyperspectral images are merged based on the similarity of each channel image to reduce hyperspectral image redundancy.

The main scene of this embodiment is: under uniform illumination, set the hyperspectral camera directly above the galvanized bridge, adjust the camera focus so that the camera field of view is the width of the bridge, collect the hyperspectral image corresponding to the bridge surface, and The hyperspectral image is processed, and the uneven galvanizing area is determined based on the characteristic information in the image, thereby realizing the monitoring of the cable tray.

Among them, the specific methods of dimensionality reduction processing are:

(1) Obtain the normalized grayscale histogram in each channel image. At this time, the horizontal axis in the grayscale histogram represents the grayscale value, and the vertical axis represents the proportion of each grayscale value;

(2) Obtain the maximum gray value and the minimum gray value in the gray histogram corresponding to each channel image, and use the proportion of each gray value in the gray range as a dimension of the vector, thus Each gray-scale histogram is converted into a vector with the same dimension as the number of gray-scale values in the gray-scale histogram, then the The grayscale features of a channel image can be expressed as/> , of which/> For the first/> The minimum gray value and the maximum gray value in each channel image,/> For the first/> The vector corresponding to the channel image;

(3) According to the channel images with the same maximum gray value and minimum gray value in each channel image, they are divided into a group. The cosine similarity between the corresponding vectors of each channel image in the same group is calculated, and the cosine similarity is calculated. The channel images corresponding to the two vectors with values greater than or equal to the similarity threshold are redundant channels for each other. In this embodiment, the similarity threshold is 0.95;

(4) Keep the channel image with the smallest dimension among the images that are redundant to each other, eliminate the other channel images, and process all the images that are redundant to each other in sequence to achieve dimensionality reduction of the hyperspectral image, and retain it after the dimensionality reduction. channel dimension sequence, and record the number of channel images after dimensionality reduction as .

It should be noted that because hyperspectral images are subdivided based on spectral bands, the spectral resolution is high and they contain many bands. Each pixel in the image corresponds to multiple channels, so that each pixel contains multiple dimensions. However, due to the strong correlation between the bands of hyperspectral images, the inter-spectral correlation coefficient of the image is large, which easily causes the stacking of hyperspectral redundant information, and this redundancy increases with the increase in the number of imaging bands and imaging resolution. , with typical high redundancy characteristics. In order to reduce the amount of calculation, it is necessary to merge the channels with greater similarity between the hyperspectral channel images to reduce the dimension of the hyperspectral image.

Step 3: Grayscale abnormality degree of each channel image.

Step 4: Hyperspectral image of uneven galvanizing.

Hyperspectral images that initially judge the existence of uneven galvanizing, because not all galvanized bridge images have uneven galvanizing, in order to avoid unnecessary operations, it is necessary to make a preliminary judgment on the uniformity of galvanizing, because when When galvanizing is uniform, the grayscale values in each channel image should be the same. Therefore, a preliminary judgment on the uniformity of galvanizing can be made based on the difference in grayscale within each channel.

Among them, the method to determine the hyperspectral image with uneven galvanizing is:

(1) Obtain the degree of grayscale abnormality of each channel image: Since the grayscale value in each channel is uniform when the galvanization is uniform, all grayscale values and their proportions in each grayscale histogram are calculated. Relative to the variance of the average gray value of the corresponding channel image, the normalized result of the obtained variance represents the degree of abnormality of the gray level in the channel image, where The grayscale abnormality degree of each channel image is recorded as/> ;

(2) Calculate the average grayscale abnormality: Since the grayscales of different metals in some channels are relatively similar, it is impossible to judge whether there are uneven galvanized areas on the surface of the bridge based on the grayscale abnormality in a single channel image only. It requires comprehensive analysis of multiple The grayscale abnormality degree of each channel image is comprehensively evaluated, so the average value of the grayscale abnormality degree of each channel image is calculated, and the average value is used as a hyperspectral image. There is a possibility of uneven galvanizing. ;

(3) According to Make a judgment: put the possibility/> and possibility threshold/> Contrast, when/> Greater than/> When , the hyperspectral image is a hyperspectral image with uneven galvanizing, and the possibility threshold in this embodiment is 0.1.

Step 5: The abnormality degree of each pixel in the channel image.

Step 6: The probability that each pixel is a galvanized abnormal pixel.

Analyze the abnormality degree of each pixel in each channel, and comprehensively determine the abnormal galvanized pixels.

Among them, the method for obtaining abnormal galvanized pixels is:

(1) Obtain the abnormality degree of each pixel in each channel image in the hyperspectral image: Since most areas in the galvanized image are still uniformly galvanized areas, the largest proportion in each channel image is The grayscale value is used as the reference grayscale value of the corresponding channel image, and the difference between the grayscale value of each pixel point and the reference grayscale value of each channel is used as the abnormality degree of the pixel point in each channel;

The calculation formula for the abnormality degree of each pixel in each channel image is:

In the formula, For the first/> The grayscale value with the largest proportion of the channel image,/> For the first/> pixel at/> Grayscale value in channel image,/> For the first/> pixel at/> The degree of abnormality in each channel image;

(2) Calculate the probability that each pixel in the hyperspectral image is a zinc-plated abnormal pixel: In order to eliminate the interference of noise points, it is necessary to combine the abnormality levels in other channel images to determine the possibility that the pixel is a galvanized abnormal point. Judgment, then the dimensionally reduced hyperspectral image The probability that a pixel is a galvanized abnormal point is/> ,Calculated as follows:

In the formula, is the number of channel images of the hyperspectral image after dimensionality reduction,/> For the first/> The probability that a pixel is an abnormal galvanized pixel;

(3) Judgment based on probability and probability threshold: when When it is less than the probability threshold, the pixels are noise points, otherwise, the /> pixels are galvanized abnormal pixels in the hyperspectral image, and the probability threshold is 0.6 in this embodiment.

It should be noted that the more channels an image has, the more likely it is to generate noise. In the same channel image, there are grayscale differences between noise and metallic zinc, but there are also grayscale differences in abnormal locations due to uneven zinc plating. Therefore, within a single channel, the suspected image obtained based on the difference between the grayscales of pixels Abnormal points include abnormal points and noise points, that is, it is impossible to distinguish abnormal points and noise points based only on the difference between gray values in a single channel image. This step takes into account that noise occurs randomly and is determined by galvanizing. The position of uniformly generated abnormal points is fixed, so it is necessary to judge the abnormal points based on the abnormality degree of the corresponding pixels in different channels.

Step 7: Optimal reference weight of each channel image.

According to the distribution of abnormal galvanized pixels, the reference weight of each channel image is adjusted to obtain the optimal reference weight for image grayscale, because when the RGB camera converts into a grayscale image, the weight of each channel is based on the human eye's sensitivity to red. Empirical values are assigned to the sensitivity of the three colors of green and blue. However, when there are more channels in the image, there is no empirical value that can be used for grayscale processing of hyperspectral images, and because the abnormality of abnormal points in each channel is not are not the same, which means that when performing grayscale processing on hyperspectral images, the weights assigned to different channels are not the same. In order to make the subsequent extraction process of uneven areas more accurate, it is often preferred to The channel image that can better highlight the abnormal area is given a greater weight. Therefore, the present invention adjusts the reference weight of the channel according to the abnormality degree of the pixels in different channels, that is, the optimal reference weight.

Among them, the optimal reference weight of each channel image is obtained as follows:

(1) Taking the grayscale abnormality degree of each channel image obtained in step 2 as the initial reference weight of the channel image, then The initial reference weight of each channel image is/> :

in, For the first/> The grayscale anomaly degree of the channel image, when/> When a channel image is fused from images with redundant channels, its grayscale abnormality is the average grayscale abnormality of each redundant channel image.

(2) Perform mean shift clustering on the coordinates of galvanized abnormal points in each channel image, where the The number of mean shift clustering results obtained in channel images is/> , each clustering result corresponds to an uneven area;

Zedi The optimal reference weight of each channel image can be expressed as:

in, For the first/> Channel image/> The number of pixels included in a clustering result can also represent the uneven area in the clustering result,/> is the total number of clustering results,/> For the first/> In the channel image, the ///> in the clustering results The abnormality degree of galvanized abnormal points in the channel image.

It should be noted that the more complete the area that can be displayed in each uneven galvanizing area in the channel image, that is, the larger the area, the greater the degree of reference for the channel image; since the purpose of the present invention is to make uneven galvanizing The simpler the area is in the subsequent segmentation process, the more obvious the uneven area needs to be. The obvious degree of the uneven area can be characterized by the abnormality of each galvanized abnormal point in the area. The greater the abnormality of the uneven area, the greater the abnormality of the uneven area. The reference degree of the channel image is also larger. Therefore, when adjusting the reference degree of the channel image in this step, it is necessary to conduct a comprehensive evaluation based on the area of the abnormal area that can be displayed in the channel image and the degree of regional abnormality.

Step 8: Hyperspectral image grayscale processing and connected domain analysis.

Step 9: Uneven galvanized areas on the surface of the galvanized cable tray.

The grayscale image corresponding to the hyperspectral image is obtained according to the reference weight of each channel, and the abnormal areas on the surface of the cable tray are extracted.

Among them, the method to obtain the uneven galvanized area is:

(1) Accumulate and sum the grayscale values of pixels in each channel image, combined with the optimal reference weight corresponding to each channel, to achieve grayscale processing of hyperspectral images;

(2) Use the Seed-Filling algorithm to analyze the connected domain of the image, and the connected area obtained is the uneven galvanized area on the surface of the cable tray.

The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the present invention. within the scope of protection.

Claims (6)

1. A method for monitoring a cable bridge based on computer vision, comprising:

collecting hyperspectral images of the surface of the cable zinc-plated bridge frame, and performing dimension reduction treatment to obtain dimension-reduced hyperspectral images;

obtaining the gray level abnormality degree of each channel image according to the difference of the gray level value of each pixel point in each channel image of the hyperspectral image after dimension reduction relative to the gray level value in the channel image;

determining hyperspectral images with zinc plating unevenness according to the gray level abnormality degree of each channel image;

obtaining the abnormal degree of each pixel point in each channel image according to the gray value of each pixel point in each channel image in the hyperspectral image with zinc plating unevenness;

calculating the probability that each pixel point is a galvanization abnormal pixel point according to the abnormality degree of the pixel point in each channel image in the hyperspectral image with galvanization non-uniformity;

determining galvanization abnormal pixel points in the hyperspectral image with galvanization non-uniformity according to the probability and the probability threshold;

performing mean shift clustering on the galvanization abnormal pixel points in each channel image of the hyperspectral image with the galvanization non-uniformity;

obtaining optimal reference weight of each channel image according to the number of pixel points in each clustering result, the abnormal degree of galvanized abnormal pixel points in the channel image and the gray level abnormal degree of the channel image;

carrying out graying treatment and connected domain analysis on the hyperspectral image according to the optimal reference weight of each channel image to obtain a zinc plating uneven area on the surface of the cable zinc plating bridge;

the method for determining the hyperspectral image with zinc plating unevenness comprises the following steps:

acquiring the gray level abnormality degree of each channel image, and obtaining a gray level abnormality degree average value as a hyperspectral image, wherein the possibility of zinc plating unevenness exists;

comparing the probability with a probability threshold, wherein when the probability is greater than the threshold, the hyperspectral image is a hyperspectral image with zinc plating unevenness;

the method for acquiring the galvanization abnormal pixel points in the hyperspectral image comprises the following steps:

obtaining the abnormal degree of each pixel point in the hyperspectral image in each channel image:

in the method, in the process of the invention,is->Gray value with maximum duty ratio of each channel image,/->Is->The pixel point is at the +.>Gray values in the individual channel images, +.>Is->The pixel point is at the +.>Degree of abnormality in the individual channel images;

calculating the probability that each pixel point in the hyperspectral image is a galvanization abnormal pixel point, wherein the calculation formula is as follows:

in the method, in the process of the invention,the number of channel images of the hyperspectral image after dimension reduction is +.>Is->The probability that each pixel point is a galvanization abnormal pixel point;

when (when)When the probability threshold is smaller than +.>The pixel points are noise points, otherwise, the first pixel point is the first pixel point>The pixel points are galvanization abnormal pixel points in the hyperspectral image; the probability threshold is set to 0.6;

the method for acquiring the optimal reference weight of each channel image comprises the following steps:

taking the gray level abnormality degree of each channel image as an initial reference weight of the channel image;

mean shift clustering coordinates of galvanized outliers in each channel image, where the firstThe number of mean shift clustering results obtained from each channel image is +.>Each cluster result corresponds to an uneven area;

the optimal reference weight calculation formula for each channel image is:

in the method, in the process of the invention,is->Optimal reference weights for the individual channel images, +.>Is->Initial reference weights of the individual channel images, +.>Is->The>The number of pixel points contained in the clustering result, namely the area of the uneven area in the clustering result,/->For the total number of clustering results, +.>Is->In the individual channel image +.>The +.>The galvanization abnormal pixel point is at the +.>Degree of abnormality in each channel image.

2. The method for monitoring the cable bridge based on computer vision according to claim 1, wherein the method for dimension reduction treatment is as follows:

acquiring a gray level histogram of each channel image normalized by the hyperspectral image;

taking the ratio of the maximum gray value and the minimum gray value in the gray histogram corresponding to each channel image and each gray value in the gray range as one dimension of a vector, and converting the gray histogram of each channel image into a vector with the same dimension as the number of the gray values in the gray histogram;

dividing channel images with the same maximum gray value and the same minimum gray value in each channel image into a group, and calculating cosine similarity between vectors corresponding to the channel images in the same group;

the channel images corresponding to the two vectors with the cosine similarity value larger than or equal to the threshold value are mutually redundant channel images, the channel image with the smallest dimension in the images which are mutually redundant channels is reserved, and other channel images are removed;

and sequentially processing all images which are mutually redundant channels, so as to realize dimension reduction of the hyperspectral image.

3. The method for monitoring the cable bridge based on computer vision according to claim 1, wherein the method for obtaining the zinc plating uneven area on the surface of the cable zinc plating bridge is as follows:

accumulating and summing gray values of pixel points in each channel image by utilizing the optimal reference weight of each channel image, and graying the hyperspectral image to obtain a gray image of the hyperspectral image;

and (3) carrying out connected domain analysis on the gray level diagram of the hyperspectral image by using a Seed-rolling algorithm, wherein the obtained connected domain is the galvanized non-uniform area on the surface of the cable galvanized bridge.

4. The method for monitoring a cable bridge according to claim 1, wherein said method for obtaining the degree of anomaly comprises:

and obtaining the abnormal degree of each pixel point in each channel image according to the difference value between the gray value of each pixel point in the hyperspectral image with zinc plating unevenness in each channel image and the gray value with the largest duty ratio in the channel image.

5. The method for monitoring a cable bridge based on computer vision according to claim 1, wherein the method for obtaining the gray level anomaly degree comprises the following steps:

and obtaining the gray level abnormality degree of each channel image according to the variance of the gray level value of each pixel point in each channel image of the hyperspectral image after dimension reduction relative to the average gray level value in the channel image.

6. The computer vision based cable bridge monitoring method according to claim 1, wherein said probability threshold is set to 0.1.