A method, device and system for visually detecting surface defects of explosion-transmitting surface sealing parts

Abstract

The present invention discloses a method, device and system for visually inspecting surface defects of explosion-transmitting surface sealing parts. The inspection method comprises the following steps: obtaining an image of the surface of the explosion-transmitting surface sealing part to be inspected; inputting the image into a trained neural network model to extract at least one defect feature area of ​​the image; and realizing defect detection according to the gray value, minimum circumscribed rectangle and roundness of the defect feature area. It can detect not only scratches and pits, but also burrs, cracks or sand holes, wrinkles or pores, and can detect many types of defects.

Description

Technical Field

[0001] The present invention belongs to the technical field of defect detection, and in particular relates to a method, device and system for visually detecting surface defects of explosion-transmitting surface sealing parts. Background Art

[0002] Defect detection usually refers to the detection of surface defects of objects, such as the detection of spots, pits, scratches, color differences, defects and other defects on the surface of workpieces. There are many existing defect detection methods. For example, the patent with application number "CN202211170480.7" discloses "a method and device for visual detection of surface defects based on linear array cameras", which uses linear array cameras and photometric stereo method to detect metal surface defects; for example, the patent with application number "CN201921568728.9" discloses "a mobile hot-rolled steel plate surface defect visual detection device", which realizes mobile visual detection of large flat metal defects.

[0003] In the development and production of detonating and ignition products, the explosive surface sealing parts are an important part of them, including transparent celluloid sheets, silk pads, and diaphragms punched from aluminum foil. The above-mentioned sheets are in bulk and disorderly piled up when they are received in the assembly process, and they need to be sorted and defect-screened before they can be used for product assembly.

[0004] Since the explosion transmission surface sealing parts include transparent celluloid sheets, silk pads and diaphragms punched from aluminum foil, the existing methods for surface defect detection have the following problems:

[0005] There are fewer types of defects that can be detected. There are various types of surface defects on the explosion-transmitting face seal parts, and the above patent only targets scratches and pits. Summary of the invention

[0006] In order to solve the problem that existing defect detection methods can detect few defect types, the present invention proposes a method, device and system for visually detecting surface defects of explosion-transmitting face sealing parts, which can realize the detection of various types of defects.

[0007] The objective of the present invention is achieved through the following technical solutions:

[0008] The first aspect of the present invention discloses a method for visually detecting surface defects of explosion transmission face sealing parts, comprising the following steps:

[0009] Obtain an image of the surface of the explosion-transmitting face sealing part to be inspected;

[0010] Inputting the image into a trained neural network model to extract at least one defect feature region of the image;

[0011] Calculating the grayscale value of the defect feature area;

[0012] In response to the gray value being less than a first threshold, calculating the area of ​​the defect feature region and the part cross-sectional area, wherein the first threshold is greater than or equal to 55 and less than or equal to 65;

[0013] In response to the area of ​​the defect characteristic region being greater than 0.0012 times the cross-sectional area of ​​the part, determining that the defect to which the defect characteristic region belongs is a wrinkle;

[0014] In response to the gray value being greater than or equal to the first threshold, determining the minimum circumscribed rectangle of the defect feature area;

[0015] Calculate the length, width and aspect ratio of the minimum circumscribed rectangle;

[0016] In response to the aspect ratio of the minimum circumscribed rectangle of the defect characteristic area being greater than a second threshold, the length of the minimum circumscribed rectangle being greater than 0.005 times the diameter of the part, and the width of the minimum circumscribed rectangle being greater than 0.2 mm, it is determined that the defect to which the defect characteristic area belongs is a scratch, a burr, or a crack, and the second threshold is greater than or equal to 2.5 and less than or equal to 3.5;

[0017] In response to the aspect ratio of the minimum circumscribed rectangle of the defect feature area being less than or equal to a second threshold, calculating the roundness of the defect feature area;

[0018] In response to the roundness of the defect feature area being greater than a third threshold, calculating the area of ​​the defect feature area and the cross-sectional area of ​​the part, wherein the third threshold is greater than or equal to 0.45 and less than or equal to 0.55;

[0019] In response to the fact that the area of ​​the defect characteristic region is greater than 0.0012 times the cross-sectional area of ​​the part, it is determined that the defect to which the defect characteristic region belongs is a sand hole, a pit or a pore.

[0020] The second aspect of the present invention discloses a device for visually inspecting surface defects of explosion-transmitting surface sealing parts, comprising a memory and a controller which are communicatively connected in sequence, wherein a computer program is stored in the memory, and the controller is used to read the computer program and execute the method for visually inspecting surface defects of explosion-transmitting surface sealing parts described in the first aspect.

[0021] The third aspect of the present invention discloses a system for visually inspecting surface defects of explosion-transmitting face sealing parts, comprising:

[0022] A parts fixing structure, wherein the parts fixing structure is used to fix the explosion transmission surface sealing parts to be tested;

[0023] An image acquisition unit, the image acquisition unit is used to acquire an image of the surface of the explosion transmission surface sealing part;

[0024] A second aspect provides a device for visually inspecting surface defects of a bursting surface sealing component, wherein the device identifies surface defects of the bursting surface sealing component based on the image.

[0025] The beneficial effects of the present invention are:

[0026] The present invention first uses a pre-trained neural network model to extract defect feature areas in an image, and then identifies irregular dark areas, long strips, and circular spot areas in the defect feature areas according to grayscale value, minimum circumscribed rectangle, and roundness, and then further determines whether the defect is a wrinkle, scratch, burr, crack, sand hole, pit, or pore based on the specific characteristics of the defect. It can identify many types of defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0028] Figure 1 It is a structural schematic diagram of a visual inspection system for surface defects of explosion transmission face sealing parts of the present invention;

[0029] Figure 2 It is a structural schematic diagram of the feeding unit;

[0030] Figure 3 It is a partial structural schematic diagram of a system for visually inspecting surface defects of explosion transmission face sealing parts of the present invention, which includes a position control component, an image acquisition unit, and an adsorption component;

[0031] Figure 4 It is a schematic structural diagram of the tray stacking bin of the present invention.

[0032] In the figure, 1-first material tray stacking bin, 2-second material tray stacking bin, 3-third material tray stacking bin, 41-third guide rail, 42-tray conveying trough, 43-second detection sensor, 44-receiving plate, 45-feeding structure, 5-position control component, 51-first guide rail, 52-second guide rail, 53-second control structure, 6-first image acquisition unit, 7-adsorption component, 8-second image acquisition unit, 9-part fixing structure, 10-waste box, 11-position detection component, 12-flexible vibration plate, 13-vibration component, 131-upper bin, 14-tray. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0034] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0036] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0037] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inside", "outside", etc. indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, or the positions or positional relationships in which the product of the invention is usually placed when in use, or the positions or positional relationships commonly understood by those skilled in the art, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0038] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" 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 a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0039] The first aspect of the present invention discloses a method for visually inspecting surface defects of explosion transmission face sealing parts, which includes steps S11 to S15. It should be noted that the step identifiers in this scheme are only for the convenience of explaining the method and do not constitute a sequence limitation. The sequence of each step is based on its language description and the sequence of each signal.

[0040] Step S11, obtaining an image of the surface of the explosion transmission surface sealing component to be detected.

[0041] The image of the surface of the explosion-transmitting face sealing part to be detected may be an image collected in real time, or an image stored locally or uploaded to the cloud.

[0042] In order to improve the accuracy of subsequent defect feature area extraction, the image can be preprocessed. The preprocessing adopts Gaussian filtering and convolution operations.

[0043] Specifically, Gaussian filtering is obtained by weighted addition of the convolution kernel, filter, and scanning window. The specific process of convolution calculation is shown in formula (1):

[0044] (1)

[0045] In the formula, is the filtering result;

[0046] is the filter function, that is, the convolution kernel The weight parameter on ;

[0047] is the pixel range value, that is And the corresponding image pixel value;

[0048] is the coordinate of the pixel in the image;

[0049] are the coordinates of the center of the convolution kernel.

[0050] Step S12: input the image into a trained neural network model to extract at least one defect feature region of the image.

[0051] The neural network model in this step needs to be pre-trained first, and the neural network model can be any one of LeNet, AlexNet, VGGNet, GoogLeNet, ResNet and YOLOv5. In order to improve the accuracy of defect feature area extraction, the YOLOv5s model in YOLOv5 is preferably used.

[0052] The YOLOv5s model consists of a backbone network layer, a feature fusion layer, and a detection layer connected sequentially.

[0053] The backbone network layer performs mosaic enhancement on the input image and randomly splices the samples by scaling, cropping, and arranging. The sample data enters the backbone network from the Focus structure, and the sample slices of 640×640×3 are spliced ​​into 320×320×12. Then, a convolution operation with 64 convolution kernels is performed to obtain the feature map, which is then completed through a series of downsampling operations such as convolution and pooling to complete the feature extraction process.

[0054] The feature fusion layer is completed through the Feature Pyramid Networks (FPN) and Perceptual Adversarial Network (PAN) structure. It is first fused with the backbone network feature map from top to bottom to achieve strong semantic feature transfer, and then achieve strong positioning feature transfer from bottom to top.

[0055] The detection layer outputs feature maps of three scales, namely 80×80, 40×40 and 20×20 grids, which are used to detect small, medium and large objects. Each grid contains three prediction boxes, each of which contains the confidence of the object and the location information of the prediction box. Finally, the repeated and redundant prediction boxes are removed through non-maximum suppression (NMS), and the prediction box information with the highest confidence is retained, thereby completing the target detection process.

[0056] When training the YOLOv5s model, a large amount of lossless and defective parts image data is used. The defects are labeled and located through a semi-supervised labeling method. Through continuous iterative learning and error correction, the network automatically selects features, and finally generates a detection model to extract defect feature areas in the image.

[0057] Step S13, calculating the grayscale value of the defect feature area; in response to the grayscale value being less than a first threshold, calculating the area of ​​the defect feature area and the cross-sectional area of ​​the part, the first threshold being greater than or equal to 55 and less than or equal to 65; in response to the area of ​​the defect feature area being greater than 0.0012 times the cross-sectional area of ​​the part, determining that the defect to which the defect feature area belongs is a wrinkle.

[0058] In this step, the irregular dark area is first screened using the grayscale value of the defect feature area. When the grayscale value is less than the first threshold, it is determined to be an irregular dark area. Then, the ratio of the area of ​​the defect feature area to the cross-sectional area of ​​the part is used to determine whether the irregular dark area is a wrinkle defect.

[0059] Specifically, the first threshold is set to any value greater than or equal to 55 and less than or equal to 65, for example, 55, 56, 60, 64, etc.

[0060] Step S14, in response to the gray value being greater than or equal to the first threshold, determining the minimum circumscribed rectangle of the defect feature area; calculating the length, width and aspect ratio of the minimum circumscribed rectangle; in response to the aspect ratio of the minimum circumscribed rectangle of the defect feature area being greater than the second threshold, the length of the minimum circumscribed rectangle being greater than 0.005 times the diameter of the part and the width of the minimum circumscribed rectangle being greater than 0.2 mm, determining that the defect to which the defect feature area belongs is a scratch, burr or crack, and the second threshold is greater than or equal to 2.5 and less than or equal to 3.5.

[0061] In this step, when the grayscale value of the defect feature area is greater than or equal to the first threshold, whether the defect feature area is a long strip is determined according to the aspect ratio of the minimum circumscribed rectangle. When the aspect ratio of the minimum circumscribed rectangle of the defect feature area is greater than the second threshold, the defect feature area is determined to be a long strip. Furthermore, when the length of the minimum circumscribed rectangle is greater than 0.005 times the diameter of the part and the width of the minimum circumscribed rectangle is greater than 0.2 mm, the defect in the defect feature area is determined to be a scratch, burr or crack.

[0062] The larger the aspect ratio, the more narrow and long it appears to the naked eye.

[0063] The second threshold is any value greater than or equal to 2.5 and less than or equal to 3.5, for example, 2.5, 2.8, 3.0, 3.1, etc.

[0064] Step S15, in response to the aspect ratio of the minimum circumscribed rectangle of the defect feature area being less than or equal to the second threshold, calculating the roundness of the defect feature area; in response to the roundness of the defect feature area being greater than the third threshold, calculating the area of ​​the defect feature area and the cross-sectional area of ​​the part, and in response to the area of ​​the defect feature area being greater than 0.0012 times the cross-sectional area of ​​the part, judging that the defect to which the defect feature area belongs is a sand hole, a pit or a pore, and the third threshold is greater than or equal to 0.45 and less than or equal to 0.55.

[0065] In this step, after judging that the defect characteristic region is neither an irregular dark region nor a long strip according to the gray value and the characteristics of the minimum circumscribed rectangle, it is judged whether it is a circular spot according to the roundness of the defect characteristic region. When the roundness of the defect characteristic region is greater than the third threshold, it is judged to be a circular spot, and further, when the area of ​​the defect characteristic region is greater than 0.0012 times the cross-sectional area of ​​the part, it is judged that the defect to which the defect characteristic region belongs is a sand hole, a pit or a pore.

[0066] Here, the roundness value is a calculation parameter used to measure the degree of proximity of the contour to a circle. A roundness value of 1 is a standard circle, and the closer it is to 0, the less close it is to a circle. For example, the roundness of a square is π / 4, which is about 0.79, and the roundness of an equilateral triangle is π*0.19, which is about 0.60.

[0067] In this step, the third threshold is any value greater than or equal to 0.45 and less than or equal to 0.55, for example, 0.45, 0.46, 0.50, 0.55, etc.

[0068] With this scheme, a pre-trained neural network model is first used to extract defect feature areas in an image, and then the irregular dark areas, long strips, and circular spots in the defect feature areas are identified in turn according to the grayscale value, the minimum circumscribed rectangle, and the roundness. Then, it is further determined whether the defect is a wrinkle, scratch, burr, crack, sand hole, pit, or pore according to the specific characteristics of the defect. There are many types of defects that can be identified, and it is suitable for defect detection of explosion-transmitting surface sealing parts of celluloid sheets made of transparent materials, cloth silk pads, and diaphragms punched out of aluminum foil.

[0069] The second aspect of the present invention discloses a visual detection device for surface defects of explosion-transmitting surface sealing parts, comprising a memory and a controller which are sequentially connected in communication, wherein a computer program is stored in the memory, and the controller is used to read the computer program and execute a visual detection method for surface defects of explosion-transmitting surface sealing parts described in the first aspect. Specifically, the memory may include, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a flash memory, a first-in-first-out memory (FIFO), or a first-in-last-out memory (FILO), etc.; the controller may not be limited to a microcontroller of the STM32F105 series.

[0070] The third aspect of the present invention discloses a system for visually inspecting surface defects of explosion-transmitting surface sealing parts. Figures 1 to 4 As shown, it includes a part fixing structure 9, an image acquisition unit and a visual detection device for surface defects of an explosion-transmitting surface sealing part described in the second aspect, wherein the part fixing structure is used to fix the explosion-transmitting surface sealing part to be detected; the image acquisition unit is used to acquire an image of the surface of the explosion-transmitting surface sealing part; and the visual detection device for surface defects of an explosion-transmitting surface sealing part identifies the surface defects of the explosion-transmitting surface sealing part according to the image.

[0071] In order to realize automatic detection, a feeding unit can be used to realize automatic loading, and there are many ways to realize it.

[0072] Since the explosion transmission surface sealing part is a circular piece, it is in the form of stacked pieces when it is tested on a large scale. In order to avoid the product being tested being multiple pieces, the feeding unit includes a flexible vibration plate 12, a loading bin 131, a vibration component 13, a position detection component 11, an adsorption component 7 and a position control component 5. Figure 2 As shown, the vibration component 13 is used to drive the upper bin 131 to vibrate so as to transport the explosion-transmitting surface sealing parts in the upper bin 131 to the flexible vibration disk 12 and drive the flexible vibration disk 12 to vibrate; the surface of the flexible vibration disk 12 in contact with the parts is flexible to achieve part protection. The position detection component is used to detect the position of the explosion-transmitting surface sealing parts on the flexible vibration disk 12. The adsorption component 7 includes a suction head and a first control structure for controlling the adsorption state of the suction head. The adsorption component 7 is an existing conventional structure and will not be described in detail here. The position control component is used to control the movement of the adsorption component 7. The visual detection device for surface defects of explosion-transmitting surface sealing parts is also signal-connected to the position detection component 11, the vibration component 13, the adsorption component 7, and the position control component.

[0073] Specifically, with the above structure, during detection, the vibration component 13 drives the flexible vibration plate 12 and the loading bin 131 to vibrate, vibrates the explosion-transmitting surface sealing parts in the loading bin 131 to the flexible vibration plate 12, and vibrates the explosion-transmitting surface sealing parts in the flexible vibration plate 12 to be placed on the flexible vibration plate 12 in a monolithic structure. The position detection component 11 detects the position of the explosion-transmitting surface sealing parts in the flexible vibration plate 12, and after the visual detection device for surface defects of the explosion-transmitting surface sealing parts receives the position information of the explosion-transmitting surface sealing parts, controls the position control component to drive the adsorption component 7 to move to the corresponding position, adsorbs the explosion-transmitting surface sealing parts and then moves to the part fixing structure, and then the suction head releases the explosion-transmitting surface sealing parts, and then the image acquisition unit acquires the image of the explosion-transmitting surface sealing parts, and the visual detection device for surface defects of the explosion-transmitting surface sealing parts performs defect detection according to the acquired image.

[0074] Since the explosion-transmitting surface sealing parts are sheet-like structures, it is necessary to conduct defect inspection on both the front and back sides. In order to achieve double-sided defect inspection and avoid flipping the parts, Figure 1 , 3 As shown, the image acquisition unit includes a first image acquisition unit 6 and a second image acquisition unit 8, and the position control component includes a first guide rail 51 for guiding the first image acquisition unit 6 and the adsorption component 7 to move back and forth in a first direction and a second control structure 53 for guiding the adsorption component 7 to move back and forth in a third direction, the first direction is placed in the horizontal direction, and the third direction is the vertical direction; the part fixing structure and the second image acquisition unit 8 are both placed below the suction head. With this structure, the first image acquisition unit 6 and the second image acquisition unit 8 are set at the same time to realize the front and back defect detection. During the detection, the first image acquisition unit 6 and the suction head can realize the defect detection of one side of the explosion transmission surface sealing part. After the surface is qualified, the suction head adsorbs the explosion transmission surface sealing part and moves it above the second image acquisition unit 8, so as to realize the defect detection of the other side. The second control structure 53 can adopt either a cylinder or an existing guide rail structure.

[0075] In order to realize automatic palletizing of trays and automatic placement of parts, the position control component also includes a second guide rail 52 for realizing the reciprocating movement and guiding of the adsorption component 7 along the second direction. The first direction and the second direction are perpendicular to each other, that is, the position control component can realize the movement control of any three-dimensional position. Its structure is an existing conventional structure, and it also includes other components, such as a cylinder for pushing the adsorption component up and down and pushing the first guide rail to move along the second guide rail. It also includes a first tray stacking bin, a second tray stacking bin, a third tray stacking bin, a feeding structure 45 for conveying trays in the first tray stacking bin to the second tray stacking bin and conveying trays in the second tray stacking bin to the third tray stacking bin, a first detection sensor for detecting whether there are trays in the first tray stacking bin, and a second detection sensor 43 for detecting whether the trays in the third tray stacking bin are full.

[0076] The first tray stacking bin is used for stacking empty trays, the second tray stacking bin is used for placing trays that are currently holding explosion-transmitting surface sealing parts, and the third tray stacking bin is used for stacking trays that are already full of explosion-transmitting surface sealing parts.

[0077] There are many structures for implementing the tray stacking bin. In order to simplify the structure and facilitate operation, such as Figure 4 As shown, it also includes a third guide rail 41 for guiding the movement of the material tray, the first material tray stacking bin, the second material tray stacking bin, and the third material tray stacking bin are sequentially arranged on the third guide rail, and a material tray conveying groove 42 is arranged between the side of the first material tray stacking bin opposite to the third material tray stacking bin and the third guide rail;

[0078] Both sides of the third material tray stacking bin are provided with a one-way rotating component that can only rotate outside the third material tray stacking bin and at least two receiving plates 44 arranged on the rotating component to realize the receiving of the material trays. The bottom of the third material tray stacking bin is provided with an upward movement control component for controlling the upward movement of the material trays in the third material tray stacking bin.

[0079] Among them, the upward movement control component can be realized by using structures such as a cylinder and a screw rod.

[0080] The tray stacking bin structure adopting this structure only requires one feeding structure. Empty trays are stacked in the first tray stacking bin 1. The feeding structure adopts structures such as pushing claws, cylinders and rails to push the trays from the first tray stacking bin 1 to the second tray stacking bin 2. After one tray is filled with qualified explosion-transmitting surface sealing parts, the feeding structure pushes the tray to the third tray stacking bin 3. At this time, the upward control component controls the tray to move upward. During the upward movement of the tray, the receiving pieces 44 on both sides are pushed to rotate. After the receiving pieces 44 rotate to a height that vacates a tray below, the upward control component retracts, and the feeding structure 45 is reset to enter the next control cycle.

[0081] The part fixing structure 9 includes a support base and a detection platform fixed on the support base. In order to facilitate the collection of defective parts, a waste box 10 can be set on one side of the part fixing structure 9, such as Figure 3 shown.

[0082] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A method for visually detecting surface defects of explosion-transmitting surface sealing parts, characterized in that: The following steps are involved: Obtain an image of the surface of the explosion-transmitting face sealing part to be inspected; Inputting the image into a trained neural network model to extract at least one defect feature region of the image; Calculating the grayscale value of the defect feature area; In response to the gray value being less than a first threshold, calculating the area of ​​the defect feature region and the part cross-sectional area, wherein the first threshold is greater than or equal to 55 and less than or equal to 65; In response to the area of ​​the defect characteristic region being greater than 0.0012 times the cross-sectional area of ​​the part, determining that the defect to which the defect characteristic region belongs is a wrinkle; In response to the gray value being greater than or equal to the first threshold, determining the minimum circumscribed rectangle of the defect feature area; Calculate the length, width and aspect ratio of the minimum circumscribed rectangle; In response to the aspect ratio of the minimum circumscribed rectangle of the defect characteristic area being greater than a second threshold, the length of the minimum circumscribed rectangle being greater than 0.005 times the diameter of the part, and the width of the minimum circumscribed rectangle being greater than 0.2 mm, it is determined that the defect to which the defect characteristic area belongs is a scratch, a burr, or a crack, and the second threshold is greater than or equal to 2.5 and less than or equal to 3.5; In response to the aspect ratio of the minimum circumscribed rectangle of the defect feature area being less than or equal to a second threshold, calculating the roundness of the defect feature area, wherein the roundness value is used to measure the degree of proximity between the contour and the circle; In response to the roundness of the defect feature region being greater than a third threshold, calculating the area of ​​the defect feature region and the cross-sectional area of ​​the part, wherein the third threshold is greater than or equal to 0.45 and less than or equal to 0.55; In response to the fact that the area of ​​the defect characteristic region is greater than 0.0012 times the cross-sectional area of ​​the part, it is determined that the defect to which the defect characteristic region belongs is a sand hole, a pit or a pore.

2. A method for visually inspecting surface defects of explosion-transmitting face sealing parts according to claim 1, characterized in that: Before inputting the image into the trained neural network model, the method further includes: Perform Gaussian filtering and convolution operations on the image.

3. A method for visually inspecting surface defects of explosion-transmitting face sealing parts according to claim 1, characterized in that: The neural network model is a YOLOv5s model.

4. A device for visually inspecting surface defects of explosion-transmitting face sealing parts, comprising a memory and a controller which are sequentially connected in communication, wherein a computer program is stored in the memory, and wherein: The controller is used to read the computer program and execute the method for visually detecting surface defects of explosion transmission face sealing parts according to any one of claims 1-3.

5. A visual inspection system for surface defects of explosion-transmitting face sealing parts, characterized in that: include: A parts fixing structure, wherein the parts fixing structure is used to fix the explosion transmission surface sealing parts to be tested; An image acquisition unit, the image acquisition unit is used to acquire an image of the surface of the explosion transmission surface sealing part; A device for visually detecting surface defects of explosion-transmitting surface sealing parts as described in claim 4, wherein the device for visually detecting surface defects of explosion-transmitting surface sealing parts identifies surface defects of explosion-transmitting surface sealing parts based on the image.

6. A visual inspection system for surface defects of explosion-transmitting face sealing parts according to claim 5, characterized in that: It also includes a feeding unit, the feeding unit comprising: Flexible vibration plate (12); Feeding bin (131); A vibration component (13), the vibration component (13) being used to drive the upper material bin (131) to vibrate so as to transport the explosion transmission surface sealing parts in the upper material bin (131) into the flexible vibration plate (12) and drive the flexible vibration plate (12) to vibrate; A position detection component (11), the position detection component being used to detect the position of a sealing part of an explosion-transmitting surface on the flexible vibration plate (12); An adsorption component (7), the adsorption component (7) comprising a suction head and a first control structure for controlling the adsorption state of the suction head; A position control component, the position control component is used to control the movement of the adsorption component (7); The explosion transmission face sealing component surface defect visual detection device is also signal-connected to the position detection component (11), the vibration component (13), the adsorption component (7), and the position control component.

7. A visual inspection system for surface defects of explosion-transmitting face sealing parts according to claim 6, characterized in that: The image acquisition unit comprises a first image acquisition unit (6) and a second image acquisition unit (8); The position control component comprises a first guide rail for guiding the first image acquisition unit (6) and the adsorption component (7) to move back and forth in a first direction, and a third guide rail for guiding the adsorption component (7) to move back and forth in a third direction, the first direction being a horizontal direction and the third direction being a vertical direction; The part fixing structure and the second image acquisition unit (8) are both placed below the suction head.

8. A visual inspection system for surface defects of explosion-transmitting face sealing parts according to claim 7, characterized in that: The position control component further comprises a second guide rail for guiding the adsorption component (7) to move back and forth along a second direction, wherein the first direction and the second direction are perpendicular to each other; It also includes a first tray stacking bin, a second tray stacking bin, a third tray stacking bin, a material conveying structure for conveying trays in the first tray stacking bin to the second tray stacking bin and conveying trays in the second tray stacking bin to the third tray stacking bin, a first detection sensor for detecting whether there are trays in the first tray stacking bin, and a second detection sensor for detecting whether the third tray stacking bin is full of trays.

9. A visual inspection system for surface defects of explosion-transmitting face sealing parts according to claim 8, characterized in that: It also includes a third guide rail for guiding the movement of the material tray, the first material tray stacking bin, the second material tray stacking bin, and the third material tray stacking bin are sequentially arranged on the third guide rail, and a material tray conveying trough is arranged between the third guide rail and the side of the first material tray stacking bin opposite to the third material tray stacking bin; Both sides of the third material tray stacking bin are provided with a one-way rotating component that can only rotate outside the third material tray stacking bin and at least two receiving plates arranged on the rotating component to realize the receiving of the material trays. The bottom of the third material tray stacking bin is provided with an upward movement control component for controlling the upward movement of the material trays in the third material tray stacking bin.

Images