Insulator defect detection device and method

Abstract

The invention relates to a glass insulator detection device. The detection device is composed of four parts, namely an objective table, a light source part, an image acquisition part and a darkroom box body part. The invention provides an insulator defect detection device and method aiming at the defects that a light refraction ring belt formed by a multi-ring-shaped groove structure of a glass insulator in an image covers scratches, cracks and the like. An insulator positioning method of multiple random circle detection and fitting degree evaluation is used to realize positioning of an insulator detection area, and the insulator detection area is divided into an umbrella ridge area and an umbrella groove area. And the positioning and segmentation of the detection area of the glass insulator are completed. Carrying out filtering fuzzification on the segmented rectangular image, carrying out difference on the image before and after filtering to obtain a residual image, and carrying out maximum between-cluster variance method segmentation on the residual image; and carrying out defect positioning identification according to the gray scale transformation characteristics of the defect area. The glass insulator defect detection is realized, and the problem that the defect detection precision is interfered due to the violent influence of the gray change of each region is solved. The novel method and technology for glass insulator defect detection are researched, and the method and technology have important significance in improving the detection efficiency, reducing the omission ratio and guaranteeing safe and stable operation of an electric power system.

Description

Technical Field

[0001] The invention relates to automatic detection technology, and in particular to a glass insulator defect detection device and method. Background Art

[0002] In high-voltage lines, the quality of insulators is related to the stability of transmission lines. The quality of insulators themselves is closely related to circuit safety. At present, several types of insulators widely used in circuits are divided into ceramic insulators, tempered glass insulators and composite insulators. Tempered glass insulators have many advantages over other types. After tempering, the mechanical strength of the surface of glass insulators is higher, making the surface structure more stable. The electrical strength of glass generally remains unchanged during the entire operation period, and its aging process is gentler than other types of insulators. A large number of self-explosions of glass insulators during operation will greatly increase the cost of power grid operation and maintenance, and seriously threaten the stability of the power grid. In view of the above problems, it is of great significance to monitor the surface defects of glass insulators to eliminate hidden dangers. In the production process of glass insulators, due to problems such as formulation, process, and ambient temperature, defects such as bubbles and cracks are prone to occur. These defects seriously affect the product quality of insulators and become a safety hazard for power grid operation.

[0003] In order to reduce such problems caused by defects in glass insulators, it is necessary to perform defect detection on glass insulators during production. Common glass parts inspection methods on production lines include hot and cold shock tests, internal water pressure tests, nickel sulfide averaging treatment and appearance inspection. The appearance inspection includes observing small defects such as scratches, bubbles, and cracks on insulators. Among the four inspection methods, the first three have been implemented on the assembly line, but the appearance inspection requires manual light inspection, which is extremely dependent on manual labor and greatly increases the workload.

[0004] The novel detection device of the present invention can detect defects of insulators quickly and effectively, and improves accuracy and efficiency. However, since the glass insulator is a slot-band structure, the bending changes of each part are large, and the reflective parts of each part are obvious, especially the junction area between the umbrella slot and the umbrella skirt. The image cannot accurately reflect the defects at the junction, resulting in inaccurate defect measurement results of the glass insulator. Summary of the invention

[0005] In order to solve the above-mentioned problem of inaccurate measurement results of glass insulators, the present invention provides a new detection device and detection method, which improves the detection efficiency.

[0006] In order to achieve the above purpose, the technical solution adopted by the present invention is as follows: the structure of the glass insulator detection device is that the detection table is an integral box-type structure, and the main body is a darkroom box to avoid interference from external light sources. The lighting source, the storage table, the light sensor, the camera and the glass insulator. The light source is placed under the transparent storage table, the glass insulator is placed above the storage table, and the camera and the light sensor are placed above the insulator and fixed to the darkroom box.

[0007] In order to obtain the ROI location of the insulator, the present invention combines the insulator prior scale and circle detection. Since the glass insulator presents a standard circular feature in the top view imaging, the circumference size of a specific type of glass insulator can be regarded as a fixed prior parameter. Based on this, the present invention proposes a preprocessing process for the input image.

[0008] In the initial stage of insulator image analysis, this study implemented a sophisticated processing operation on the input insulator image. Specifically, the operation aims to convert the original image from the RGB (red, green, blue) three-channel color space to a grayscale image. This conversion step significantly simplifies the complexity of the image data, by removing the color information and retaining only the brightness information, thus providing a more concise data basis for subsequent data processing and analysis. This conversion not only helps to reduce the computational complexity, but also speeds up the detection process and improves the overall processing efficiency. The specific implementation method of this preprocessing process is shown in formula (1), which realizes the conversion of color space through mathematical transformation.

[0009] P gray (i,j)=0.299R+0.587G+0.114B (1)

[0010] Where P gray It represents the gray value of the pixel after grayscale conversion, while R, G, and B represent the channel components of each color channel respectively.

[0011] Based on the grayscale processed image, the present invention further adopts the Hough circle detection algorithm to achieve accurate positioning of the insulator target. Hough transform is widely used in the field of image processing due to its strong and robust detection ability for circular structures. Through this algorithm, the circular features presented by the insulator can be effectively identified from the grayscale image, thereby determining the specific position of the insulator. At this stage, based on the previously located insulator position, we define the grayscale image area where the insulator is located, and clarify the ROI area for subsequent detection and analysis. This process is of great significance for improving detection accuracy, reducing calculation amount, and optimizing algorithm performance.

[0012] Furthermore, by dividing the processed image into regions, each region is separated, and different detections are performed on different regions to avoid the grayscale changes of the interlaced bands affecting the detection results.

[0013] Furthermore, the polar coordinates of the images of each region are expanded, and the detection region is transformed from a circular ring to a rectangle to facilitate subsequent defect detection.

[0014] Furthermore, the radially expanded grayscale image of the insulator consists of the following parts: noise in the image, non-defective parts with slight grayscale value changes in the image, and suspected defective parts with drastic grayscale value changes. By fuzzifying the image, the noise can be effectively filtered out, and the drastic change parts can be made smooth, but the background area with small grayscale gradient changes and the non-defective area of ​​the umbrella groove are slightly affected. Therefore, the residual is obtained by differentiating the images before and after smoothing, and the noise in the residual image is basically eliminated. For the obtained insulator target image, the slowly changing part corresponds to the smooth background of the umbrella groove area of ​​the image, and the drastic change part corresponds to the suspected defective area of ​​the umbrella groove area. According to the above theory and image characteristics, the present invention adopts a new glass insulator defect detection method. First, the input insulator image is blurred by smoothing mean filtering; then, the image before and after blurring is differentiating to obtain the residual, and the residual image is segmented by the maximum inter-class variance method; finally, the segmented binary image is subjected to regional connectivity detection and global regional threshold defect recognition combined to perform defect location recognition.

[0015] Furthermore, the suspected defect area is obtained by the maximum between-class variance method (OTSU) to locate the defect area. This detection method can not only avoid the problem of insufficient detection accuracy caused by the interlacing of the slot band area, but also accurately identify the defect location of the insulator. The detection result is more accurate and the efficiency is higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of an insulator detection device and method of the present invention;

[0017] Figure 2 It is a schematic front view of a glass insulator defect detection device and method thereof in a specific embodiment 1 of the present invention;

[0018] Figure 3 It is a top view schematic diagram of a glass insulator defect detection device and method in a specific embodiment 1 of the present invention;

[0019] Reference numerals: 1, darkroom box; 2, camera; 3, storage table; 4, LED light source; 5, glass insulator DETAILED DESCRIPTION

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

[0021] like Figure 1The figure shows a flow chart of an insulator detection device and method of the present invention, which comprises the following steps:

[0022] Step S1: Building a testing platform. In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0023] Step S2: Collect insulator image data through a camera, take a photo of the insulator for identification, and input a target image.

[0024] Step S3: Convert the original image from the RGB (red, green, blue) three-channel color space to a grayscale image. This conversion step significantly simplifies the complexity of the image data by removing color information and retaining only brightness information, thereby providing a more concise data basis for subsequent data processing and analysis.

[0025] Step S4: Identify the circular features of the insulator from the grayscale image, so as to determine the specific position of the insulator. Further, by dividing the processed image into regions, the regions are separated and different detections are performed on different regions.

[0026] Step S5: By blurring the image, the noise can be effectively filtered out, and the drastically changing parts can be made smooth, but the background area with small grayscale gradient changes and the defect-free area of ​​the umbrella groove are slightly affected. Therefore, the residual is obtained by differentiating the image before and after smoothing, and the noise in the residual image is basically eliminated. For the obtained insulator target image, the slowly changing part corresponds to the smooth background of the umbrella groove area of ​​the image, and the drastically changing part corresponds to the suspected defect area of ​​the umbrella groove area. According to the above theory and image characteristics, the present invention adopts a new glass insulator defect detection method. First, the input insulator image is blurred by smoothing mean filtering; then, the image before and after blurring is differentiating to obtain the residual, and the residual image is segmented by the maximum inter-class variance method.

[0027] Step S6: Obtain suspected defective areas through the maximum between-class variance method (OTSU) to locate the defective areas. This detection method can perfectly avoid problems such as insufficient detection accuracy caused by interweaving of insulator umbrella slot areas, and can accurately identify the location of the defective area.

[0028] Embodiment 1;

[0029] Reference Figure 2 , 3As shown, a glass insulator defect detection device and method include a darkroom box, a transparent storage table, a camera and an LED light source. The LED light source is placed under the storage table to facilitate the light source to evenly illuminate the insulator through the storage table. This lighting method can fully and evenly illuminate the insulator to be inspected, and back lighting can better highlight defects.

[0030] Light source is an important part of the whole detection system. Choosing the right light source can greatly improve the imaging quality of the camera, making the details of each part of the image clearer, thus solving the problem of subsequent processing difficulties caused by unclear imaging. In addition, a good light source should weaken the background and highlight the object to a certain extent, providing convenience for subsequent image processing.

[0031] Reference Figure 2 , 3 As shown, an insulator defect detection device and method, the camera is placed just above the transparent storage table, which is convenient for comprehensive, uniform illumination target image acquisition, and the target image is output by acquiring the uniform image, and the image is subsequently processed for defect identification.

[0032] The image after basic processing is used to define the grayscale image area where the insulator is located, and to clarify the ROI area for subsequent detection and analysis. This process is helpful to improve detection accuracy, reduce calculation amount and optimize algorithm performance.

[0033] The image after regional positioning and recognition is segmented and expanded in polar coordinates. Since the insulator after positioning is in a circular shape, it is not convenient for subsequent processing, so the insulator is transformed from a circular shape to a rectangular shape. The expanded insulator image becomes a rectangle, which is convenient for subsequent image processing and defect recognition.

[0034] The grayscale image of the insulator after positioning and radial expansion consists of the following parts: noise in the image, non-defective parts with slight grayscale value changes in the image, and suspected defective parts with drastic grayscale value changes. By fuzzifying the image, the noise can be effectively filtered out and the drastic change parts can be smoothed, but the background area with small grayscale gradient changes and non-defective areas have little effect. Therefore, the residual is obtained by differentiating the images before and after smoothing, and the noise in the residual image is basically eliminated. For the obtained insulator target image, the slowly changing part corresponds to the smooth background of the image area, and the drastic change part corresponds to the suspected defective area of ​​the area. According to the above theory and image characteristics, the present invention adopts a new glass insulator defect detection device and method. First, the input insulator image is blurred by smoothing mean filtering; then, the image before and after blurring is differentiating to obtain the residual, and the residual image is segmented by the maximum inter-class variance method; finally, the segmented binary image is subjected to regional connectivity detection and global regional threshold defect recognition combined to perform defect location and recognition.

[0035] The present invention can effectively detect the defects of various parts of the insulator through the setting of the detection device and the detection method, avoids the problem of inaccurate detection results caused by the interaction of various areas, greatly reduces the situation of false detection and missed detection, and makes the measurement result more accurate and efficient.

[0036] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings, but the above is only a preferred embodiment of the present invention. It should be noted that the technical features involved in the above embodiments and their preferred methods can be reasonably combined and matched into a variety of equivalent schemes by those skilled in the art without departing from or changing the design ideas and technical effects of the present invention; therefore, the present invention is not limited to the scope of implementation shown in the drawings, and all changes made in accordance with the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the specification and drawings, should be within the protection scope of the present invention.

Claims

1. A glass insulator detection device and method, characterized in that: The structure of the glass insulator detection device is that the detection table is an integral box-type structure, and the main body is a darkroom box to avoid interference from external light sources. The lighting source, the storage table, the light sensor, the camera and the glass insulator are placed under the transparent storage table, the glass insulator is placed above the storage table, and the camera and the light sensor are placed above the insulator and fixed to the darkroom box.

2. The insulator detection device according to claim 1, characterized in that: The box body is a steel pipe and a black sunshade.

3. The insulator detection device according to claim 1, characterized in that: The light source is an LED planar light source.

4. The insulator detection device according to claim 1, characterized in that: The storage platform is a ring-shaped transparent storage platform.

5. The insulator detection device according to claim 1, characterized in that: The camera is a CCD industrial camera.

6. The insulator detection device according to claim 1, characterized in that: The side of the box is connected to a computer for subsequent defect detection and identification processes.

7. The insulator detection device according to claim 6, characterized in that: The following steps are involved: Step S1: Use the camera of claim 5 to take a picture to obtain the overall image of the insulator, and transmit the picture information to the computer as described in claim 6. Step S2: Perform image preprocessing on the image acquired in step S1 to reduce the image space occupancy and facilitate subsequent image processing. Step S3: Position the ROI region of the image in step S2 to identify the part of the insulator that needs to be inspected. Since the insulator is divided into slot band structures, this step divides the slot band structure into different slot band structures. Step S4: Expand the structure described in step S3 in polar coordinates to expand the annular structure into a rectangular structure to facilitate subsequent defect identification. Step S5: Perform defect detection on the content described in step S4. The grayscale image of the insulator after positioning in step S3 and radial expansion in step S4 consists of the following parts: noise in the image, non-defective parts with slight changes in grayscale values ​​in the image, and suspected defective parts with drastic changes in grayscale values. By blurring the image, the noise can be effectively filtered out and the drastic changes can be made smooth, but the background area with small changes in grayscale gradient and the non-defective area of ​​the umbrella groove are slightly affected. Therefore, the residual is obtained by differentiating the images before and after smoothing, and the noise in the residual image is basically eliminated. For the obtained insulator target image, the slowly changing part corresponds to the smooth background of the umbrella groove area of ​​the image, and the drastic changes correspond to the suspected defective area of ​​the umbrella groove area. According to the aforementioned theory and image characteristics, this paper adopts a new glass insulator defect detection method. First, the input insulator image is blurred by smoothing mean filtering; Subsequently, the image difference before and after blurring is used to obtain the residual, and the residual image is segmented using the maximum inter-class variance method. Finally, the segmented binary image is subjected to regional connectivity detection and global regional threshold defect recognition for defect location and recognition.

Images