Industrial defective product detection method and defective product detection equipment

Abstract

The invention belongs to the technical field of visual inspection, and particularly relates to an industrial defective product detection method and defective product detection device.The industrial defective product detection method comprises the steps of S1, image collection, S2, image preprocessing, after image collection, image preprocessing to improve image quality, S3, key features in images are extracted through a computer vision algorithm, and S4, defect recognition and classification. S5, decision making and judgment: classifying the products through a processing system according to the identified defects or abnormal characteristics, and judging whether the products are defective products, S6, outputting results, judging whether the products are qualified or unqualified, and finally outputting qualified or unqualified judgment results, and the system is connected with a production line, automatically sorts the defective products or prompts an operator to carry out manual intervention. According to an image sensor, a light source, image processing software and an analysis algorithm, the product appearance and defects are recognized and classified, and through the image obtaining and processing technology, whether the product meets the design standard or not is detected in real time, and the problems of surface defects, shape inconformity, size out-of-tolerance and the like are recognized.

Description

Technical Field

[0001] The present invention relates to the technical field of visual inspection of industrial defective products, and specifically to a method for detecting industrial defective products and a device for detecting defective products. Background Art

[0002] During the manufacturing process of industrial components, in order to ensure the qualified rate of the products of industrial engineering components, all products need to be inspected to avoid defective phenomena such as hole cracks, large holes or foreign objects, which may cause the inability to achieve the preset functions after installing components on subsequent industrial engineering components and make rework difficult.

[0003] During the inspection process of traditional industrial engineering components, they are all observed by the naked eye of workers, and defective products are marked after being found. This inspection method not only has a large manual labor intensity and low inspection efficiency, but also the inspection results cannot be uploaded to the system in a timely manner and associated with other information of the products. Therefore, it will affect the accuracy and timeliness of tracing defective products in the later stage. Summary of the Invention

[0004] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Simplifications or omissions may be made in this part, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this part, the abstract and the title, and such simplifications or omissions shall not be used to limit the scope of the present invention.

[0005] Therefore, the purpose of the present invention is to provide a method for detecting industrial defective products and a device for detecting defective products, which can realize the recognition and classification of the appearance and defects of products based on an image sensor (such as a CCD or CMOS camera), a light source, image processing software, and an analysis algorithm. Through image acquisition and processing technology, it can detect in real time whether the product meets the design standards and identify problems such as surface defects, non-conforming shapes, and dimensional tolerances. To solve the above technical problems, according to one aspect of the present invention, the following technical solutions are provided:

[0006] A method for detecting industrial defective products, which includes the following steps:

[0007] S1. Image acquisition;

[0008] Camera selection: Select a suitable camera according to the size, precision requirements, and lighting conditions of the detection object. Such as industrial CCD cameras, CMOS cameras;

[0009] Lens and light source configuration: Configure a suitable lens and light source according to the surface characteristics, shape, and size of the detection object;

[0010] S2. Image preprocessing;

[0011] After image acquisition, preprocess the image to improve the image quality for subsequent processing;

[0012] Grayscale conversion: Convert the color image to a grayscale image to simplify calculations;

[0013] Denoising: Remove the noise in the image through filtering techniques such as median filtering and Gaussian filtering;

[0014] Enhance contrast: Increase the contrast of the image through histogram equalization or other image enhancement methods to make the defects more obvious;

[0015] S3. Feature extraction;

[0016] Extract the key features in the image through computer vision algorithms;

[0017] S4. Defect identification and classification;

[0018] Template matching: Use a known standard template to match the current image to determine whether the shape of the object is compliant for the detection of surface defects and non-compliant geometric shapes;

[0019] Machine learning and deep learning: Use a convolutional neural network (CNN) deep learning model to classify the image and identify defects. By training the model, identify different types of defects such as cracks, bubbles, scratches, etc., and calibrate whether it is a defective product;

[0020] S5. Decision-making and determination;

[0021] Based on the identified defects or abnormal features, the processing system will classify the product and determine whether it is a defective product;

[0022] S6. Output results;

[0023] Qualified / Unqualified determination: Finally, output the judgment result of qualified or unqualified. The system is connected to the production line to automatically sort defective products or prompt the operator for manual intervention;

[0024] Data recording and feedback: Record the detection data and defect information to provide data support for quality improvement, and upload it to the remote cloud disk. Through the feedback mechanism, optimize the production process.

[0025] As a preferred solution of an industrial defective product detection method described in the present invention, wherein: in the step S1, the commonly used light sources include ring light sources, backlight sources, and directional light sources to enhance the contrast of surface defects of the object.

[0026] As a preferred solution of an industrial defective product detection method described in the present invention, wherein: in the step S3, the key feature extraction includes the following:

[0027] Edge detection: By using edge detection algorithms such as the Sobel operator and the Canny algorithm, the edges of objects in the image are detected to further determine the contours and defects of the objects;

[0028] Texture analysis: Detect whether there are abnormalities in the texture of the object surface, such as surface defects like scratches and cracks;

[0029] Morphological processing: Processes such as erosion and dilation can help identify whether the shape, size, etc. of the product meet the requirements.

[0030] As a preferred solution of an industrial defective product detection method described in the present invention, wherein: in the step S5, common decision methods include:

[0031] Threshold determination: The system compares the numerical values of features such as size and color contrast with the set standard threshold;

[0032] Defect area analysis: The system determines the proportion of the defect in the surface or volume of the product. If it exceeds the set threshold, it is regarded as a defective product;

[0033] Statistical analysis: The system conducts statistical analysis on the data of multiple detections and performs standardized detection through indicators such as the mean and variance.

[0034] As a preferred solution of an industrial defective product detection method described in the present invention, wherein: the step S6 further includes adding a detection mark to the component message that has completed the safety detection. Subsequent devices can quickly determine whether the message has completed the safety detection by judging whether there is a detection mark, which helps to improve the detection efficiency.

[0035] An industrial defective product detection device, characterized in that it includes:

[0036] An operating table frame serving as an operating table board, and multiple connecting frames are connected to the top of the operating table frame;

[0037] A feeding component, connected to the operating table frame, and conveys the industrial components to be detected to the detection component for detection;

[0038] A detection component, placed on the operating table frame, includes a support frame connected to the central position at the top of the operating table frame. An expansion motor is connected to the outside of the support frame, the expansion end of the expansion motor is connected to a sliding frame, a mounting frame is connected to the sliding frame, a rotating seat is arranged on the outside of the mounting frame, and a detection camera is connected to the movable end of the rotating seat.

[0039] As a preferred embodiment of an industrial defective product detection device according to the present invention, the following is provided: The connecting frames are symmetrically arranged along the top of the operation table frame. A feeding telescopic cylinder is connected to the top of the left connecting frame, and a feeding robotic gripper is connected to the output end of the feeding telescopic cylinder. A picking telescopic cylinder is connected to the top of the right connecting frame, and a picking robotic arm is connected to the output end of the picking telescopic cylinder.

[0040] As a preferred embodiment of an industrial defective product detection device according to the present invention, the following is provided: The feeding component includes a conveying frame connected to the top of the operation table frame. A conveyor belt is connected to the conveying frame, and the conveyor belt penetrates through both connecting frames.

[0041] As a preferred embodiment of an industrial defective product detection device according to the present invention, the following is provided: The feeding telescopic cylinder, the feeding robotic gripper, the picking telescopic cylinder, the picking robotic arm, the feeding component, and the detection component are all connected to the remote control terminal.

[0042] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0043] 1. The vision-based defective product detection method realizes the automated detection of various problems such as surface defects, dimensional deviations, and shape non-conformities of products through advanced image processing technologies and intelligent algorithms, and has the advantages of high efficiency, accuracy, and strong repeatability.

[0044] 2. Through data recording and feedback, the detection data and defect information are recorded, providing data support for quality improvement and uploading to the remote cloud disk. Through the feedback mechanism, the production process is optimized. At the same time, when outputting products, according to the final judgment result of qualified or unqualified, the system connects to the production line to automatically sort defective products or prompt the operator for manual intervention to ensure the sorting efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the present invention will be described in detail below with reference to the drawings and detailed embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts. Among them:

[0046] Figure 1 It is a schematic diagram of the steps of the detection method of the present invention;

[0047] Figure 2 It is a schematic diagram of the overall structure of the present invention;

[0048] Figure 3 It is a schematic diagram of a partial structure of the present invention.

[0049] In the figure: 100 operating console frame, 110 connecting frame, 120 feeding telescopic cylinder, 121 feeding mechanical claw, 130 material taking telescopic cylinder, 131 material taking robotic arm, 200 feeding component, 210 conveying frame, 220 conveyor belt, 300 detection component, 310 support frame, 320 telescopic motor, 321 sliding frame, 322 mounting frame, 330 rotating base, 331 detection camera. Detailed implementation manners

[0050] In order to make the above objects, features and advantages of the present invention more obvious and understandable, the following will describe in detail the specific implementation manners of the present invention with reference to the accompanying drawings.

[0051] In the following description, many specific details are set forth in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific implementation manners disclosed below.

[0052] Secondly, the present invention will be described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally out of the general scale, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width and depth should be included.

[0053] In order to make the purpose, technical solution and advantages of the present invention clearer, the following will further describe in detail the embodiments of the present invention with reference to the accompanying drawings.

[0054] The present invention provides an industrial defective product detection method and a defective product detection device. By relying on an image sensor (such as a CCD or CMOS camera), a light source, image processing software, and an analysis algorithm, the identification and classification of the appearance and defects of products are realized. Through image acquisition and processing technologies, it is detected in real time whether the product meets the design standards, and problems such as surface defects, shape non-conformities, and dimensional tolerances are identified. Please refer to Figures 1-3 , including the following steps:

[0055] Please continue to refer to Figure 1 , S1, Image acquisition;

[0056] Camera selection: Select a suitable camera according to the size, precision requirements, and lighting conditions of the detection object. Such as industrial CCD cameras, CMOS cameras;

[0057] Lens and light source configuration: Configure a suitable lens and light source according to the surface characteristics, shape, and size of the detection object;

[0058] S2, Image preprocessing;

[0059] After image acquisition, preprocess the image to improve its quality for subsequent processing;

[0060] Grayscale conversion: Convert the color image to a grayscale image to simplify calculations;

[0061] Denosing: Remove noise in the image through filtering techniques such as median filtering and Gaussian filtering;

[0062] Enhance contrast: Increase the contrast of the image through histogram equalization or other image enhancement methods to make defects more obvious;

[0063] S3. Feature extraction;

[0064] Extract key features in the image through computer vision algorithms;

[0065] S4. Defect recognition and classification;

[0066] Template matching: Use a known standard template to match the current image to determine whether the shape of the object is compliant for detecting surface defects and non - compliant geometric shapes;

[0067] Machine learning and deep learning: Use a convolutional neural network (CNN) deep learning model to classify the image and identify defects. Through training the model, identify different types of defects such as cracks, bubbles, scratches, etc., and determine whether it is a defective product;

[0068] S5. Decision - making and determination;

[0069] Based on the identified defects or abnormal features, the processing system will classify the product and determine whether it is a defective product;

[0070] S6. Output results;

[0071] Qualified / unqualified determination: Finally, output the judgment result of qualified or unqualified. The system is connected to the production line to automatically sort defective products or prompt the operator for manual intervention;

[0072] Data recording and feedback: Record the detection data and defect information to provide data support for quality improvement and upload it to the remote cloud disk. Through the feedback mechanism, optimize the production process;

[0073] Furthermore, in step S1, commonly used light sources include ring light sources, backlight sources, and directional light sources to enhance the contrast of surface defects of the object;

[0074] Furthermore, in step S3, key feature extraction includes the following:

[0075] Edge detection: Detect the edges of objects in the image through edge detection algorithms such as the Sobel operator and the Canny algorithm to further judge the contour and defects of the objects;

[0076] Texture analysis: Detect whether there are abnormalities in the texture of the object surface, such as surface defects like scratches and cracks;

[0077] Morphological processing: Such as corrosion, dilation, etc., which can help identify whether the shape, size, etc. of the product meet the requirements;

[0078] Furthermore, in step S5, common decision-making methods include:

[0079] Threshold determination: The system compares the numerical values of features, such as size and color contrast, with the set standard threshold;

[0080] Defect area analysis: The system determines the proportion of the defect in the surface or volume of the product. If it exceeds the set threshold, it is regarded as a defective product;

[0081] Statistical analysis: The system conducts statistical analysis on the data of multiple detections and performs standardized detection through indicators such as mean and variance;

[0082] Furthermore, step S6 also includes adding a detection mark to the component message that has completed the safety detection. Subsequently, the subsequent device can quickly determine whether the message has completed the safety detection by judging whether there is a detection mark on the message, which helps to improve the detection efficiency;

[0083] Please continue to refer to Figures 2-3 , an industrial defective product detection device, characterized by including:

[0084] An operating table frame 100 serving as an operating table frame board, and a plurality of connecting frames 110 are connected to the top of the operating table frame 100;

[0085] The connecting frames 110 are symmetrically arranged along the top of the operating table frame 100. A feeding telescopic cylinder 120 is threadedly connected to the top of the left connecting frame 110. The feeding telescopic cylinder 120 is used to drive the feeding mechanical claw 121 to place the component to be detected on the conveyor belt 220. The output end of the feeding telescopic cylinder 120 is screwed with the feeding mechanical claw 121. A picking telescopic cylinder 130 is threadedly connected to the top of the right connecting frame 110. The output end of the picking telescopic cylinder 130 is screwed with the picking robotic arm 131. The picking telescopic cylinder 130 is used to drive the picking robotic arm 131 to take out and classify the detected components. By setting the feeding telescopic cylinder 120 and the picking telescopic cylinder 130, the feeding mechanical claw 121 and the picking robotic arm 131 are driven to act. The feeding mechanical claw 121 places the component to be detected on the conveyor belt 220, and the picking robotic arm 131 takes out the detected component, realizing an integrated operation of feeding, detection, and picking;

[0086] Please continue to refer to Figures 2-3, the feeding component 200 is connected to the operation table frame 100, and conveys the industrial components to be detected to the detection component 300 for detection;

[0087] The feeding component 200 includes a conveying frame 210 connected to the top of the operation table frame 100. A conveyor belt 220 is connected to the conveying frame 210, and the conveyor belt 220 penetrates through the two ends of the connecting frame 110;

[0088] Please continue to refer to Figures 2-3 , the detection component 300 is placed on the operation table frame 100, and includes a support frame 310 threadedly connected to the center of the top of the operation table frame 100. A telescopic motor 320 is screwed on the outside of the support frame 310. The telescopic end of the telescopic motor 320 is threadedly connected to a sliding frame 321. An installation frame 322 is threadedly connected to the sliding frame 321. A rotating seat 330 is threadedly connected to the outside of the installation frame 322. A detection camera 331 is connected to the movable end of the rotating seat 330;

[0089] Action:

[0090] The telescopic motor 320 works, drives the detection camera 331 placed on the installation frame 322 to move, and cooperates with the rotating seat 330 to realize the detection of different positions of the component by the detection camera 331;

[0091] Furthermore, the feeding telescopic cylinder 120, the feeding mechanical claw 121, the picking telescopic cylinder 130, the picking robotic arm 131, the feeding component 200 and the detection component 300 are all connected to the remote control terminal;

[0092] Through steps S1 - S6, based on the image sensor (such as a CCD or CMOS camera), light source, image processing software, and analysis algorithm, to realize the recognition and classification of the product appearance and defects. Through image acquisition and processing technology, it can detect in real time whether the product meets the design standards, identify problems such as surface defects, shape non - conformity, and dimensional tolerance. At the same time, through data recording and feedback, it records the detection data and defect information, provides data support for quality improvement, and uploads it to the remote cloud disk. Through the feedback mechanism, it optimizes the production process. And when outputting the product, according to the final judgment result of qualified or unqualified, the system connects to the production line, automatically sorts out defective products or prompts the operator for manual intervention to ensure the sorting efficiency;

[0093] Although the present invention has been described above with reference to the embodiments, various modifications can be made thereto and components thereof can be replaced with equivalents without departing from the scope of the present invention. In particular, as long as there is no structural conflict, the features in the embodiments disclosed by the present invention can be combined with each other in any way, and the exhaustive description of these combinations is not given in this specification only for the consideration of saving space and resources. Therefore, the present invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for detecting industrial defective products, characterized in that: The steps include: S1, image acquisition; Camera selection: Choose a suitable camera based on the size of the object to be inspected, accuracy requirements, and lighting conditions, such as industrial CCD cameras and CMOS cameras; Lens and light source configuration: configure appropriate lens and light source according to the surface characteristics, shape and size of the detected object; S2, image preprocessing; After image acquisition, image preprocessing is performed to improve image quality and facilitate subsequent processing; Grayscale: Convert color images to grayscale images to simplify calculations; Denoising: remove noise from the image through filtering techniques, such as median filtering and Gaussian filtering; Enhance contrast: Increase the contrast of the image through histogram equalization or other image enhancement methods to make defects more obvious; S3, feature extraction; Extract key features from images through computer vision algorithms; S4. Defect identification and classification; Template matching: Use a known standard template to match the current image to determine whether the shape of the object is compliant, as a detection of surface defects and geometric shapes that do not meet the standards; Machine learning and deep learning: Use the convolutional neural network (CNN) deep learning model to classify images and identify defects. By training the model, different types of defects such as cracks, bubbles, scratches, etc. can be identified and marked as defective or not. S5. Decision-making and judgment; Based on the identified defects or abnormal features, the processing system will classify the product to determine whether it is defective; S6, output results; Pass / fail judgment: The final pass or fail judgment result is output. The system connects to the production line and automatically sorts out defective products or prompts the operator to intervene manually. Data recording and feedback: Record inspection data and defect information to provide data support for quality improvement, upload to a remote cloud disk, and optimize the production process through a feedback mechanism.

2. The method for detecting industrial defective products according to claim 1, characterized in that: In the step S1, the light sources commonly used include an annular light source, a backlight source, and a directional light source to enhance the contrast of surface defects of the object.

3. The method for detecting industrial defective products according to claim 2, characterized in that: In step S3, key feature extraction includes the following: Edge detection: Use edge detection algorithms, such as the Sobel operator and the Canny algorithm, to detect the edges of objects in the image to further determine the contours and defects of the objects; Texture analysis: Detecting whether there are any abnormalities in the texture of the object surface, such as scratches, cracks and other surface defects; Morphological processing: such as corrosion, expansion and other processing, can help identify whether the shape, size, etc. of the product meet the requirements.

4. The method for detecting industrial defective products according to claim 3, characterized in that: In step S5, common decision-making methods include: Threshold determination: The system compares the numerical value of features, such as size and color contrast, with the set standard threshold; Defect area analysis: The system determines the proportion of defects to the surface or volume of the product. If it exceeds the set threshold, it is considered a defective product; Statistical analysis: The system performs statistical analysis on data from multiple tests and conducts standardized testing through indicators such as mean and variance.

5. The method for detecting industrial defective products according to claim 4, characterized in that: The step S6 also includes adding a detection mark to the component message that has completed the safety detection. The subsequent device determines whether the message has the detection mark to quickly determine whether the message has completed the safety detection, which helps to improve the detection efficiency.

6. An industrial defective product detection device, characterized in that: include: An operating platform (100) serving as an operating platform panel, wherein a plurality of connecting frames (110) are connected to the top of the operating platform (100); A feeding component (200) is connected to the operating platform (100) and conveys the industrial component to be inspected to the inspection component (300) for inspection; The detection component (300) is placed on the operating platform (100), and comprises a support frame (310) connected to the central position of the top of the operating platform (100), the outer side of the support frame (310) is connected to a telescopic motor (320), the telescopic end of the telescopic motor (320) is connected to a sliding frame (321), the sliding frame (321) is connected to a mounting frame (322), a rotating seat (330) is arranged on the outer side of the mounting frame (322), and the movable end of the rotating seat (330) is connected to a detection camera (331).

7. The industrial defective product detection equipment according to claim 6, characterized in that: A plurality of groups of the connecting frames (110) are symmetrically arranged along the top of the operating platform (100); the top of the left connecting frame (110) is connected to a feeding telescopic cylinder (120); the output end of the feeding telescopic cylinder (120) is connected to a feeding mechanical claw (121); the top of the right connecting frame (110) is connected to a material picking telescopic cylinder (130); the output end of the material picking telescopic cylinder (130) is connected to a material picking mechanical arm (131).

8. The industrial defective product detection equipment according to claim 7, characterized in that: The feeding component (200) comprises a conveying frame (210) connected to the top of the operating platform (100), a conveying belt (220) is connected to the conveying frame (210), and the conveying belt (220) passes through the connecting frames (110) at both ends.

9. The industrial defective product detection equipment according to claim 8, characterized in that: The feeding telescopic cylinder (120), the feeding mechanical claw (121), the material picking telescopic cylinder (130), the material picking mechanical arm (131), the feeding component (200) and the detection component (300) are all connected to the remote control end.

Images