CN111209876B - Oil leakage defect detection method and system - Google Patents

Abstract

The invention provides a method and a system for detecting oil leakage defects, wherein the system comprises the following steps: the device comprises an ultraviolet light source module, a visible light imaging module, a parameter setting module, a display module, an alarm module and an image analysis module; the method comprises the following steps: s1: fixing the relative positions of an ultraviolet light source module and a visible light imaging module, wherein the irradiation direction of the ultraviolet light source module is relatively parallel to the shooting direction of the visible light imaging module, and after the ultraviolet light source module irradiates the area to be detected, the visible light imaging module collects high-definition image data of the area to be detected and transmits the high-definition image data to an image analysis module; s2: the image analysis module receives the high-definition image data, and an oil leakage defect detection and identification result is obtained through image analysis processing; s3: and transmitting the image and the character information of the oil leakage defect detection identification result to a display module for displaying, and transmitting the detection identification result signal to an alarm module for alarming. The invention enhances the anti-interference capability of the environment and avoids the influence of personal experience on the detection result.

Description

Oil leakage defect detection method and system

Technical Field

The present invention relates to the technical field of oil leakage detection and identification, and more specifically, to an oil leakage defect detection method and system.

Background Art

At present, most methods for detecting oil leaks use manual inspections, which require certain experience and are identified by "visual inspection" and "sniffing". The effect of manual inspections is greatly affected by work experience, and it takes a lot of time and is inefficient. In recent years, fluorescence detection technology has been widely used in the fields of oil exploration and development, monitoring of oil pollutant content in soil, and monitoring of aircraft hydraulic oil leakage and sea surface oil spills, and has obtained a large number of research results, which have verified the effectiveness of oil fluorescence detection methods. However, there are problems such as lack of anti-interference ability to the environment and high requirements for image acquisition equipment.

Chinese patent CN108844689A proposes a transformer oil leakage detection method. This patent uses the fluorescent characteristics of transformer oil components, combined with substation inspections, and for different oil-immersed equipment, first conducts a large-area rough search to observe whether there is a blue fluorescence phenomenon, and then accurately locates the area of oil blue fluorescence. According to the depth and range of blue fluorescence, the location of oil leakage and the severity of leakage can be determined. This method requires manual inspections, and requires the inspectors to have certain experience in interference with other fluorescent substances to determine the oil leakage. The detection results are affected by personal subjective consciousness and experience.

Chinese patent CN110455463A proposes a system and method for detecting oil pollution in substation equipment. This patent uses the characteristic that oil pollution will produce fluorescence under the irradiation of ultraviolet laser, and determines the fluorescent area in the photo taken by the ultraviolet fluorescent camera as the oil pollution area. This method must ensure that there are no other fluorescent interferences in the environment, otherwise false alarms will occur and lack the ability to resist environmental interference.

Chinese patent CN110174220A proposes a transformer on-load tap changer oil leakage detection system and method. The patent sends ultraviolet light to the transformer on-load tap changer to be detected, then receives and processes the fluorescence radiated by the transformer on-load tap changer to be detected after receiving the ultraviolet light, and converts the fluorescence into a digital signal through a photoelectric signal sensor, and finally analyzes the received digital signal through a data analysis device to obtain the oil leakage of the transformer on-load tap changer to be detected. This method requires filtering the fluorescence into monochromatic light through a monochromatic light filter, and then converting it into a digital signal for analysis, which has high requirements for image acquisition equipment.

At present, the main defects of oil leakage defect detection and identification are as follows:

(1) The test results are affected by personal subjective consciousness and experience. In the inspection site, other fluorescent substances are generally present. When exposed to ultraviolet light, non-oil fluorescent substances will also emit fluorescence, which will interfere with the judgment of oil leakage. The inspection personnel need to have certain experience to judge and eliminate non-oil fluorescent substances in order to determine the oil leakage. This process and test results are affected by personal subjective consciousness and experience, resulting in the lack of stability of the test results.

(2) Lack of anti-interference ability to the environment. The fluorescent area in the photo taken by the ultraviolet fluorescent camera is determined as the oil pollution area. However, when other non-oil fluorescent substances appear in a complex environment, false alarms will occur.

(3) High requirements for image acquisition equipment. It is necessary to design new equipment or modify existing equipment, such as adding monochromatic light filters, to acquire images and then perform fluorescent substance detection and analysis to detect oil leaks.

Summary of the invention

The primary purpose of the present invention is to provide an oil leakage defect detection method, enhance the anti-interference ability of the environment, avoid the influence of personal subjective consciousness, experience, etc. on the detection results, and reduce the requirements for image acquisition equipment.

A further object of the present invention is to provide an oil leakage defect detection and identification system.

In order to solve the above technical problems, the technical solution of the present invention is as follows:

A method for detecting oil leakage defects comprises the following steps:

S1: Fix the relative positions of the ultraviolet light source module and the visible light imaging module. The irradiation direction module of the ultraviolet light source module and the shooting direction of the visible light imaging module are both facing the area to be detected and the two directions are relatively parallel. The relative parallelism means that a certain deflection angle is allowed in the two directions. To ensure the ultraviolet irradiation effect and the high-definition image acquisition effect, the deflection angle range is 0°-10°. The visible light imaging module parameters are set using the parameter setting module. After the ultraviolet light source module irradiates the area to be detected, the visible light imaging module collects high-definition image data of the area to be detected and transmits the high-definition image data to the image analysis module.

S2: The image analysis module receives high-definition image data and obtains the oil leakage defect detection and recognition results through image analysis processing;

S3: The oil leakage defect detection and identification result is transmitted to the display module for display, the oil leakage defect detection and identification result signal is transmitted to the alarm module for alarm, and the parameters of the display module and the alarm module are set using the parameter setting module.

Preferably, the visible light imaging module is a high-definition camera.

Preferably, in step S1, the oil at the oil leak in the area to be detected will emit fluorescence due to the fluorescence effect under the irradiation of the ultraviolet light source, and the visible light imaging module collects a high-definition fluorescence image.

Preferably, the image analysis processing steps of the image analysis module in step S2 are specifically as follows:

S21: using the color and brightness characteristics of the oil fluorescence in the high-definition fluorescence image, the fluorescence area is extracted as the suspected oil leakage area through channel transformation and threshold processing. If there is no fluorescence area, it is determined that there is no oil leakage defect, the captured high-definition fluorescence image is output, and a no oil leakage signal is issued; if there is a fluorescence area, it goes to step S22;

S22: Analyze the area characteristics of the suspected oil leakage area. There are often small areas of fluorescent substances that interfere with the detection in the area to be detected, such as plastic debris, residual oil stains, etc., which appear as small fluorescent areas under ultraviolet light. Use area screening to remove small fluorescent areas and eliminate the influence of small areas of fluorescent substances that interfere with the detection;

S23: Performing shape feature analysis on the suspected oil leakage area determined by area feature analysis, including eccentricity and rectangularity analysis; if no suspected oil leakage area meets the shape feature analysis, it is determined that there is no oil leakage, and the captured high-definition fluorescent image is output, and a no oil leakage signal is issued; if there is a suspected oil leakage area that meets the shape feature analysis, the suspected oil leakage area continues to be judged and identified, and enters step S24;

S24: Convert the original high-definition fluorescence multi-channel image into a single-channel grayscale image;

S25: performing grayscale feature analysis of the suspected oil leakage area judged and screened by the area feature and the shape feature in the single-channel grayscale image, including grayscale mean analysis; judging the suspected oil leakage area by the grayscale feature, the grayscale mean is the grayscale mean of the grayscale value set of the grayscale image corresponding to the suspected oil leakage area, reflecting the brightness of the area. The oil leakage area is generally brighter and has a larger grayscale mean;

S26: Based on the judgment and recognition results of the comprehensive area characteristics, shape characteristics, and grayscale characteristics, the oil leakage situation of the area to be detected is determined; after the characteristic analysis and judgment of each suspected oil leakage area, if an oil leakage area exists, it is determined that there is an oil leakage defect, and a processed image with the oil leakage area marked is output, and an oil leakage alarm signal is issued; if there is no oil leakage area, it is determined that there is no oil leakage defect, the captured high-definition fluorescent image is output, and a no oil leakage signal is issued.

Preferably, extracting the fluorescent area in step S21 specifically includes the following steps:

S211: Preprocessing the input high-definition fluorescence image, wherein the preprocessing includes denoising and exponential transformation. In order to remove noise while maintaining image edge detail information as much as possible, adaptive median filtering is used for denoising. In order to further highlight the fluorescence area in the high-definition image, exponential transformation is used to enhance the contrast of the high gray value area.

S212: The high-definition fluorescence image is an RGB three-channel image, and the high-definition fluorescence image is decomposed into an R channel image, a G channel image, and a B channel image;

S213: using the R channel image, the G channel image, and the B channel image to convert and obtain an H (hue) channel image and a V (value) channel image in the HSV color model;

S214: region 1 is obtained by threshold processing the H (hue) channel image, and region 1 is a region extracted according to the fluorescence color feature. The color characteristics of the oil fluorescence are used to recognize the oil fluorescence color in the H (hue) channel image through threshold processing, thereby realizing the extraction of the region with the fluorescence color; region 2 is obtained by threshold processing the V (brightness) channel image, and region 2 is a region extracted according to the fluorescence brightness feature. The brightness characteristics of the oil fluorescence are used to recognize the oil fluorescence brightness in the V (brightness) channel image through threshold processing, thereby realizing the extraction of the region with the fluorescence brightness;

S215: Performing an intersection operation on region 1 and region 2 to obtain an intersection region, which is the extracted fluorescence region.

Preferably, in step S213, the conversion relationship of converting the R channel, G channel, and B channel images into the H (hue) channel image and the V (brightness) channel image is:

R'＝R/255

G'＝G/255

B'＝B/255

C max = max(R', G', B')

Cmin＝mim(R'，G'，B')

Δ＝C max-C min

V＝Cmax

Where R, G, and B are the R channel image, G channel image, and B channel image respectively, R’, G’, and B’ are the normalized images of the R channel image, G channel image, and B channel image respectively, H is the H (hue) channel image in the HSV color model, and V is the V (lightness) channel image in the HSV color model.

Preferably, the shape feature analysis in step S23 includes eccentricity analysis. The eccentricity can reflect the stretching degree of the region, which is specifically expressed as calculating the inertia principal axis ratio. The calculation formula is as follows:

为平均向量，μ_ij为各阶中心矩，e为偏心度计算结果，待检测区常存在尼龙扎带等起固定作用的塑料制品，均表现为细且长的特性，偏心度一般较大。In the above formula, R is the regional point set, n is the number of point sets, x is the horizontal coordinate of the point set, y is the vertical coordinate of the point set, S is the area of the region,

is the average vector, μ _ij is the central moment of each order, and e is the eccentricity calculation result. In the inspection area, there are often plastic products such as nylon ties that play a fixing role. They are thin and long, and the eccentricity is generally large.

Preferably, the shape feature analysis in step S23 includes rectangularity analysis, where the rectangularity represents the degree of similarity between an object and a rectangle, and the calculation formula is as follows:

In the above formula, _AS is the area of the region, _AR is the minimum rectangular area surrounding the region. There are often interfering fluorescent substances such as rectangular labels and stickers in the area to be detected, and the rectangularity is generally close to 1. Calculating the rectangularity can remove the interfering fluorescent substances in the rectangle.

Preferably, the conversion relationship for converting the original high-definition fluorescent multi-channel image into a single-channel grayscale image in step S24 is:

Gray(i,j)=0.299*R(i,j)+0.578*G(i,j)+0.114*B(i,j)

Where (i, j) is the pixel coordinate, R, G, and B are the R channel image, G channel image, and B channel image respectively.

Preferably, the oil leakage situation of the area to be tested is determined in step S26, specifically:

If the suspected oil leakage area meets the area characteristics, shape characteristics and grayscale characteristics, it is determined that the suspected oil leakage area has leaked oil; if the suspected steam leakage area does not meet one or more of the area characteristics, shape characteristics and grayscale characteristics, it is ruled out that the suspected oil leakage area has leaked oil.

A system using the above-mentioned oil leakage defect detection method comprises:

Image acquisition subsystem and information interaction subsystem;

The image acquisition subsystem includes an ultraviolet light source module and a visible light imaging module;

The information interaction subsystem includes a parameter setting module, a display module, an alarm module, and an image analysis module;

The ultraviolet light source module emits ultraviolet light to illuminate the area to be detected;

The visible light imaging module collects high-definition visible light images of the area to be detected, and transmits the high-definition image data to the image analysis module;

The parameter setting module is used by the user to set visible light imaging module parameters, result display settings, and alarm information settings;

The display module displays the oil leakage defect detection and identification result image and related information prompts;

The alarm module sends an alarm prompt to the outside world according to the oil leakage defect detection and identification result;

The image analysis module receives high-definition image data, and transmits the image text information and alarm signal of the oil leakage defect detection and identification result obtained through image analysis processing to the display module and the alarm module respectively.

Depending on the number of monitoring points and the complexity of the environment, inspection robots, fixed pan-tilt heads, and handheld devices can be used as image acquisition system platforms as needed.

Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

The present invention irradiates the inspection area with ultraviolet light to visually display the oil at the oil leakage point in the form of blue or purple fluorescence, and uses a high-definition camera to capture the image of the area without adding hardware such as monochromatic light filters to increase the requirements for image acquisition equipment. The digital image processing method is used to judge the area, shape and grayscale characteristics of the suspected oil leakage area, eliminates the influence of other non-oil fluorescent substances on the detection results, avoids the influence of personal subjective consciousness, experience, etc. on the detection results, eliminates the influence of other non-oil fluorescent substances on the detection results, and ensures the accuracy and stability of the detection results.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic flow chart of the method of the present invention.

FIG. 2 is a schematic diagram of the image analysis processing steps of the image analysis module.

FIG3 is a schematic diagram of the process of extracting the fluorescence area.

FIG. 4 is a schematic diagram of a system of the present invention.

DETAILED DESCRIPTION

The drawings are for illustrative purposes only and are not to be construed as limiting the present patent;

In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

Example 1

This embodiment provides an oil leakage defect detection method, as shown in FIG1 , which specifically includes the following steps:

S1: Fix the relative positions of the ultraviolet light source module and the visible light imaging module, the irradiation direction of the ultraviolet light source and the shooting direction of the high-definition camera are both facing the area to be detected, and the two directions are relatively parallel, wherein the parameters of the visible light imaging module are set by using the parameter setting module, after the ultraviolet light source module irradiates the area to be detected, the visible light imaging module collects high-definition image data of the area to be detected and transmits the high-definition image data to the image analysis module;

S2: The image analysis module receives high-definition image data and obtains the oil leakage defect detection and recognition results through image analysis processing;

S3: The image and text information of the oil leakage defect detection and identification result are transmitted to the display module for display, the oil leakage defect detection and identification result signal is transmitted to the alarm module for alarm, and the parameters of the display module and the alarm module are set using the parameter setting module.

Furthermore, the relative parallelism described in S1 allows for a certain deviation angle. To ensure the ultraviolet irradiation effect and high-definition image acquisition effect, the deviation angle range is preferably less than 10°.

Furthermore, the oil at the leaking site will fluoresce due to the fluorescence effect under the irradiation of ultraviolet laser, and the visible light imaging module will collect high-definition fluorescence images.

As shown in FIG2 , the image analysis process S2 of the image analysis module specifically includes:

S21: Using the high-definition fluorescent image and the color and brightness characteristics of the oil fluorescence, the fluorescent area is extracted as the suspected oil leakage area through channel transformation and threshold processing. If there is no fluorescent area, it is determined that there is no oil leakage defect, and the captured high-definition fluorescent image is output, and a no oil leakage signal is issued; if there is a fluorescent area, the oil leakage judgment and identification will continue.

As shown in FIG3 , extracting the fluorescence area specifically includes:

S211: Preprocessing the input high-definition fluorescence image, including: denoising and exponential transformation. In order to remove noise while maintaining the image edge detail information as much as possible, adaptive median filtering is used for denoising; in order to further highlight the fluorescence area in the high-definition image, exponential transformation is used to enhance the contrast of the high gray value area.

S212: Decompose the high-definition fluorescent RGB three-channel image into an R channel image, a G channel image, and a B channel image.

S213: Use the R channel, G channel, and B channel image conversion to obtain the H (hue) channel image and the V (value) channel image in the HSV color model.

Furthermore, the conversion relationship between the R channel, G channel, and B channel images into the H (hue) channel image and the V (brightness) channel image is:

R'＝R/255

G'＝G/255

B'＝B/255

C max = max(R', G', B')

Cmin＝min(R'，G'，B')

Δ＝C max-C min

V＝Cmax

Where R, G, and B are the R channel image, G channel image, and B channel image respectively, R’, G’, and B’ are the normalized images of the R channel image, G channel image, and B channel image respectively, H is the H (hue) channel image in the HSV color model, and V is the V (lightness) channel image in the HSV color model.

S214: Obtain region 1 by threshold processing on the H (hue) channel image; obtain region 2 by threshold processing on the V (value) channel image.

Furthermore, by utilizing the color characteristics of oil fluorescence, the oil fluorescence color is recognized through threshold processing in the H (hue) channel image, thereby realizing the extraction of the fluorescent color area, with a lower limit of 170° and an upper limit of 225°.

Furthermore, by utilizing the brightness characteristics of oil fluorescence, the oil fluorescence brightness is recognized through threshold processing in the V (brightness) channel image, thereby realizing the extraction of the fluorescence brightness area, with the lower limit value being 0.8 and the upper limit value being 0.98.

S215: Performing an intersection operation on region 1 and region 2 to obtain an intersection region, which is the extracted fluorescence region.

S22: Analyze the area characteristics of the suspected oil leak area. There are often small areas of fluorescent substances that interfere with detection in the area to be detected, such as plastic debris, residual oil stains, etc., which actually appear as small fluorescent areas with an area of less than 1 square centimeter under ultraviolet light. Use area screening to remove small areas and eliminate the influence of small areas of fluorescent substances that interfere with detection.

S23: Perform shape feature analysis on the suspected oil leakage area judged by area feature, including eccentricity and rectangularity analysis; if no suspected oil leakage area meets the shape feature analysis, it is determined that there is no oil leakage, the captured high-definition fluorescent image is output, and a no oil leakage signal is issued; if there is a suspected oil leakage area that meets the shape feature analysis, the suspected oil leakage area continues to be judged and identified, and enters step S24.

Furthermore, the eccentricity can reflect the stretching degree of the region. Specifically, it is expressed as calculating the inertia principal axis ratio, and the calculation formula is as follows:

为平均向量，μ_ij各阶中心矩，e为偏心度计算结果。待检测区常存在尼龙扎带等起固定作用的塑料制品，均表现为细且长的特性，偏心度一般大于7，漏油区域偏心度一般小于4。In the formula, R is the regional point set, n is the number of point sets, x is the horizontal coordinate of the point set, y is the vertical coordinate of the point set, S is the area of the region,

is the average vector, μ _ij is the central moment of each order, and e is the eccentricity calculation result. The area to be tested often has plastic products such as nylon ties that play a fixing role, which are thin and long, and the eccentricity is generally greater than 7. The eccentricity of the oil leakage area is generally less than 4.

Furthermore, the rectangularity indicates the similarity between an object and a rectangle, reflecting the degree to which the object fills its circumscribed rectangle, and the calculation formula is as follows:

In the formula, _AS is the area of the region, and _AR is the minimum rectangular area surrounding the region. The area to be detected often contains interfering fluorescent substances such as rectangular labels and stickers, and the rectangularity is generally greater than 0.9, while the rectangularity of the oil leakage area is generally less than 0.6.

S24: Convert the original input high-definition fluorescence multi-channel image into a single-channel grayscale image.

Furthermore, the conversion relationship from high-definition image to grayscale image is:

Gray(i,j)=0.299*R(i,j)+0.578*G(i,j)+0.114*B(i,j)

Where (i, j) is the pixel coordinate, R, G, and B are the R channel image, G channel image, and B channel image respectively.

S25: For the suspected oil leakage area judged and screened by area and shape characteristics, grayscale characteristics of the area are analyzed in the grayscale image, including grayscale mean analysis; the suspected oil leakage area is judged by the grayscale characteristics.

Furthermore, the grayscale mean is the grayscale mean of the grayscale value set of the grayscale image corresponding to the suspected oil leakage area, reflecting the brightness of the area. The grayscale mean of the oil leakage area can have a lower limit of 150 and an upper limit of 255.

S26: Determine the oil leakage situation in the area to be tested based on the judgment and recognition results of the comprehensive area characteristics, shape characteristics, and grayscale characteristics. If the suspected oil leakage area meets the area characteristics, shape characteristics, and grayscale characteristics, it is determined that the suspected oil leakage area has oil leakage; if the suspected steam leakage area does not meet one or more of the area characteristics, shape characteristics, and grayscale characteristics, it is ruled out that the suspected oil leakage area has oil leakage. After analyzing and judging the characteristics of each suspected oil leakage area, if there is an oil leakage area, it is determined that there is an oil leakage defect, and the processed image with the oil leakage area marked is output, and an oil leakage alarm signal is issued; if there is no oil leakage area, it is determined that there is no oil leakage defect, and the captured high-definition fluorescent image is output, and a no oil leakage signal is issued.

Example 2

This embodiment provides an oil leakage defect detection and identification system, as shown in FIG4 , including an image acquisition subsystem and an information interaction subsystem;

The image acquisition subsystem includes an ultraviolet light source module and a visible light imaging module;

The information interaction subsystem includes a parameter setting module, a display module, an alarm module, and an image analysis module;

The ultraviolet light source module emits ultraviolet light to illuminate the area to be detected;

The visible light imaging module collects high-definition visible light images of the area to be detected, and transmits the high-definition image data to the image analysis module;

The parameter setting module is used by the user to set visible light imaging module parameters, result display settings, and alarm information settings;

The display module displays the oil leakage defect detection and identification result image and related information prompts;

The alarm module sends an alarm prompt to the outside world according to the oil leakage defect detection and identification result;

The image analysis module receives high-definition image data, and transmits the image text information and alarm signal of the oil leakage defect detection and identification result obtained through image analysis processing to the display module and the alarm module respectively.

The same or similar reference numerals correspond to the same or similar components;

The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims (8)

1. A method for detecting oil leakage defects, characterized in that it comprises the following steps: S1: Fix the relative positions of the ultraviolet light source module and the visible light imaging module, wherein the irradiation direction of the ultraviolet light source module and the shooting direction of the visible light imaging module are both facing the area to be detected and the two directions are relatively parallel, wherein the parameters of the visible light imaging module are set using the parameter setting module, and after the ultraviolet light source module irradiates the area to be detected, the visible light imaging module collects high-definition image data of the area to be detected and transmits the high-definition image data to the image analysis module; S2: The image analysis module receives high-definition image data and obtains the oil leakage defect detection and recognition results through image analysis processing; S3: passing the oil leakage defect detection and identification result to the display module for display, passing the oil leakage defect detection and identification result signal to the alarm module for alarm, and setting the parameters of the display module and the alarm module by using the parameter setting module; The image analysis processing steps of the image analysis module in step S2 are specifically as follows: S21: using the color and brightness characteristics of the oil fluorescence in the high-definition fluorescence image, the fluorescence area is extracted as the suspected oil leakage area through channel transformation and threshold processing. If there is no fluorescence area, it is determined that there is no oil leakage defect, the captured high-definition fluorescence image is output, and a no oil leakage signal is issued; if there is a fluorescence area, it goes to step S22; S22: Analyze the area characteristics of the suspected oil leak area and use area screening to remove small fluorescent areas that interfere with detection; S23: Performing shape feature analysis on the suspected oil leakage area determined by area feature analysis, including eccentricity and rectangularity analysis; if no suspected oil leakage area meets the shape feature analysis, it is determined that there is no oil leakage, and the captured high-definition fluorescent image is output, and a no oil leakage signal is issued; if there is a suspected oil leakage area that meets the shape feature analysis, the suspected oil leakage area continues to be judged and identified, and enters step S24; S24: Convert the original high-definition fluorescence multi-channel image into a single-channel grayscale image; S25: performing grayscale feature analysis of the suspected oil leakage area judged and screened by area features and shape features in the single-channel grayscale image, including grayscale mean analysis; judging the suspected oil leakage area by grayscale features; S26: Based on the judgment and recognition results of the comprehensive area features, shape features, and grayscale features, the oil leakage situation of the area to be detected is determined; after the characteristics of each suspected oil leakage area are analyzed and judged, if there is an oil leakage area, it is determined that there is an oil leakage defect, and a processed image with the oil leakage area marked is output, and an oil leakage alarm signal is issued; if there is no oil leakage area, it is determined that there is no oil leakage defect, and the high-definition fluorescent image taken is output, and a no oil leakage signal is issued; The shape feature analysis in step S23 includes eccentricity analysis, which is specifically expressed as calculating the inertia principal axis ratio, and the calculation formula is as follows:

In the above formula, R is the regional point set, n is the number of point sets, x is the horizontal coordinate of the point set, y is the vertical coordinate of the point set, S is the area of the region,

is the average vector, μ _ij is the central moment of each order, and e is the calculated result of eccentricity. 2. The oil leakage defect detection method according to claim 1 is characterized in that, in step S1, the oil at the oil leakage site in the area to be detected will emit fluorescence due to the fluorescence effect under the irradiation of the ultraviolet light source, and the visible light imaging module collects a high-definition fluorescence image. 3. The oil leakage defect detection method according to claim 1 is characterized in that extracting the fluorescent area in step S21 specifically comprises the following steps: S211: preprocessing the input high-definition fluorescence image, wherein the preprocessing includes denoising and exponential transformation; S212: The high-definition fluorescence image is an RGB three-channel image, and the high-definition fluorescence image is decomposed into an R channel image, a G channel image, and a B channel image; S213: using the R channel image, the G channel image, and the B channel image to convert and obtain an H (hue) channel image and a V (value) channel image in the HSV color model; S214: performing threshold processing on the H (hue) channel image to obtain region 1, where region 1 is a region extracted based on the fluorescence color feature; performing threshold processing on the V (brightness) channel image to obtain region 2, where region 2 is a region extracted based on the fluorescence brightness feature; S215: Performing an intersection operation on region 1 and region 2 to obtain an intersection region, which is the extracted fluorescence region. 4. The oil leakage defect detection method according to claim 3 is characterized in that the conversion relationship of converting the R channel, G channel, and B channel images into the H channel image and the V channel image in step S213 is: R′＝R/255 G′＝G/255 B′＝B/255 C max=max(R′,G′,B′) C min=min(R′, G′, B′) Δ＝C max-C min

V＝Cmax Where R, G, and B are the R channel image, G channel image, and B channel image respectively, R’, G’, and B’ are the normalized images of the R channel image, G channel image, and B channel image respectively, H is the H (hue) channel image in the HSV color model, and V is the V (lightness) channel image in the HSV color model. 5. The oil leakage defect detection method according to claim 4 is characterized in that the shape feature analysis in step S23 includes rectangularity analysis, and the rectangularity represents the degree of similarity between an object and a rectangle, and the calculation formula is as follows:

In the above formula, _AS is the area of the region, and _AR is the area of the minimum rectangle enclosing the region. 6. The oil leakage defect detection method according to claim 5 is characterized in that the conversion relationship of converting the original high-definition fluorescent multi-channel image into a single-channel grayscale image in step S24 is: Gray(i,j)=0.299*R(i,j)+0.578*G(i,j)+0.114*B(i,j) Where (i, j) is the pixel coordinate, R, G, and B are the R channel image, G channel image, and B channel image respectively. 7. The oil leakage defect detection method according to claim 6 is characterized in that the oil leakage situation of the area to be tested is determined in step S26, specifically: If the suspected oil leakage area meets the area characteristics, shape characteristics and grayscale characteristics, it is determined that the suspected oil leakage area has leaked oil; if the suspected steam leakage area does not meet one or more of the area characteristics, shape characteristics and grayscale characteristics, it is ruled out that the suspected oil leakage area has leaked oil. 8. A system using the oil leakage defect detection method according to any one of claims 1 to 7, characterized in that it comprises: An ultraviolet light source module, wherein the ultraviolet light source module emits ultraviolet light to illuminate the area to be detected; A visible light imaging module, which collects high-definition visible light images of the area to be detected and transmits the high-definition image data to the image analysis module; A parameter setting module, which is used by the user to set visible light imaging module parameters, result display settings, and alarm information settings; A display module, which displays images of oil leakage defect detection and identification results and prompts related information; An alarm module, which sends an alarm prompt to the outside world according to the oil leakage defect detection and identification result; The image analysis module receives high-definition image data, and transmits the image text information and alarm signal of the oil leakage defect detection and identification result obtained through image analysis processing to the display module and the alarm module respectively.

Images