Automobile part detection system and method based on image recognition

Abstract

The invention discloses an automobile part detection system and method based on image recognition. The system comprises an image acquisition assembly, a preprocessing assembly, a recognition assembly, a data analysis assembly and a control assembly, the image acquisition assembly is used for acquiring an image of a to-be-detected automobile part, and the image acquisition assembly communicates with the preprocessing assembly through a data channel; the pre-processing assembly is used for pre-processing the received image and then transmitting the image to the identification assembly; the recognition assembly processes the received image data based on a preset recognition model and generates recognition data of the automobile parts, the data analysis assembly is used for generating a quality evaluation report of the automobile parts based on the recognition data, and the control assembly is used for adjusting the working state of the production line according to the quality evaluation report. The whole structure is simple, through cooperation of the image acquisition assembly and the preprocessing assembly, the image of the to-be-detected automobile part can be rapidly acquired and processed, manual operation steps are reduced, and the whole efficiency of a production line is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of automobiles, and in particular to an automobile parts detection system and method based on image recognition. Background Art

[0002] Automobile parts inspection is an important part of the automobile manufacturing process. Various parameters of automobile parts are checked through detection devices to ensure that the parts meet performance requirements during assembly and operation. Traditional automobile parts inspection systems mainly rely on manual inspection through basic mechanical tools. Workers usually visually inspect the surface quality and shape integrity of parts. The defect of this method is that it is affected by the workers' experience, skills and work intensity, leading to misjudgment or omission. In addition, workers also usually use mechanical measuring tools such as gauges, calipers and micrometers to detect the size and tolerance of parts. Although this method has high accuracy, it is slow and suitable for measuring simple shapes but not for detecting complex surface defects. The above method cannot meet the current automated production line. Therefore, in order to solve the above problems, an automobile parts inspection system and method based on image recognition is needed. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention proposes an automobile parts detection system and method based on image recognition. Among them, the specific technical solution of an automobile parts detection system based on image recognition is as follows:

[0004] An automobile parts detection system based on image recognition, characterized in that:

[0005] It includes image acquisition component, preprocessing component, recognition component, data analysis component and control component;

[0006] The image acquisition component is used to acquire the image of the automobile part to be detected, and the image acquisition component communicates with the pre-processing component through a data channel;

[0007] The preprocessing component is used to preprocess the received image and transmit it to the recognition component;

[0008] The recognition component processes the received image data based on a preset recognition model and generates recognition data of automobile parts. The data analysis component is used to generate a quality assessment report of the automobile parts based on the recognition data. The control component is used to adjust the working state of the production line according to the quality assessment report.

[0009] In order to better realize the present invention, it is possible to further:

[0010] The preprocessing component is used to perform grayscale conversion, denoising and edge detection on the received image.

[0011] Further: the identification data includes a part category identifier, a matching score and a defect status.

[0012] A working method of an automobile parts detection system based on image recognition The specific technical scheme is as follows:

[0013] A working method of an automobile parts detection system based on image recognition, characterized in that:

[0014] S1: The image acquisition component acquires the image of the automobile part to be detected and transmits the image to the preprocessing component;

[0015] S2: the preprocessing component preprocesses the received automobile part image and inputs the binary image data into the recognition component;

[0016] S3: the recognition component performs feature extraction and feature matching on the binary image data based on a preset image recognition model, and the recognition component obtains recognition data through feature matching;

[0017] S4: the identification component sends the identification data to the data analysis component;

[0018] S5: After receiving the identification data from the identification component, the data analysis component analyzes the identification data and generates an evaluation report for the automobile parts;

[0019] S6: The data analysis component sends the evaluation report to the control component;

[0020] S7: The control component automatically adjusts the working state of the detection system according to the evaluation report.

[0021] In order to better realize the present invention, it is possible to further:

[0022] The S2 comprises the following steps:

[0023] S2-1: The preprocessing component performs denoising on the grayscale image through the grayscale algorithm of the cv2.cvtColor function;

[0024] S2-2: The preprocessing component eliminates noise in the image through a bilateral filtering denoising algorithm;

[0025] S2-3: The preprocessing component performs edge detection operation through the edge detection algorithm of the Plaus operator to extract edge information of the part and generate an edge image;

[0026] S2-4: the preprocessing component converts the edge image into a binary edge image matrix B′;

[0027] Generate a binary edge image matrix B′:

[0028]

[0029] Among them, B′(x,y)=0 is the edge of the part;

[0030] B′(x,y)=1 is the background area;

[0031] S2-5: The preprocessing group inputs the edge image matrix B′ into the recognition component.

[0032] In order to better realize the present invention, it is possible to further:

[0033] S3-1: The recognition component is provided with a contour detection model C, which extracts key shape features of the edge image matrix B′ to obtain an extraction vector F:

[0034] Where F = C(B′)

[0035] F=[f1,f2,...,f n ]

[0036] f i is the part feature parameter;

[0037] S3-2: A standard part feature database is provided, the identification component is connected to the standard part feature database, and the standard part feature database has a standard vector group {S1, S2, ..., S m}, where S i is a preset standard feature vector in the standard vector group;

[0038] S3-3: The recognition component sequentially compares the extracted vector F with the standard vector group {S1, S2, ..., S m Each S in i By comparison, we can calculate F and each S in the vector group according to the following formula: i Similarity:

[0039] d(F,S i )=‖FS i ‖

[0040] S3-4: The recognition component obtains the best matching feature S c , according to the optimal matching feature S c , determine the state of the part and generate identification data R:

[0041] R = [id, sc, ds]

[0042] Among them, id is the identifier of the part category, sc is the similarity score, and ds is the defect type of the part.

[0043] The beneficial effects of the present invention are as follows: the overall structure is simple, and by setting the cooperation of the image acquisition component and the preprocessing component, the image of the automobile parts to be inspected can be quickly acquired and processed, the manual operation steps can be reduced, and the overall efficiency of the production line can be greatly improved. By setting the preprocessing component, the preprocessing component performs optimization processing such as graying, denoising and edge detection on the image to ensure high-quality input of image data. The recognition component relies on a preset image recognition model, and through precise feature extraction and matching technology, it can accurately locate the features and defects of the parts, realize high-precision recognition of complex parts, and reduce the possibility of missed detection and false detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a system diagram of the present invention;

[0045] Figure 2 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. 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.

[0047] like Figure 1 As shown:

[0048] An automobile parts detection system based on image recognition includes an image acquisition component, a preprocessing component, a recognition component, a data analysis component and a control component;

[0049] The image acquisition component is used to acquire the image of the automobile part to be detected, and the image acquisition component communicates with the pre-processing component through a data channel;

[0050] The preprocessing component is used to preprocess the received image and transmit it to the recognition component, wherein the preprocessing component is used to grayscale, remove noise and perform edge detection on the received image.

[0051] The recognition component processes the received image data based on a preset recognition model and generates recognition data of the automobile part, specifically, the recognition data includes a part category identifier, a matching score, and a defect status. The data analysis component is used to generate a quality assessment report of the automobile part based on the recognition data, and the control component is used to adjust the working state of the production line according to the quality assessment report.

[0052] like Figure 2As shown, a specific technical scheme of a working method of an automobile parts detection system based on image recognition is as follows:

[0053] A working method of an automobile part detection system based on image recognition comprises the following steps: S1: the image acquisition component acquires an image of an automobile part to be detected, and transmits the image to a preprocessing component;

[0054] S2: The preprocessing component preprocesses the received automobile part image and inputs the binary image data into the recognition component.

[0055] Specifically: S2 includes the following steps:

[0056] S2-1: The preprocessing component performs denoising on the grayscale image through the grayscale algorithm of the cv2.cvtColor function;

[0057] S2-2: The preprocessing component eliminates noise in the image through a bilateral filtering denoising algorithm;

[0058] S2-3: The preprocessing component performs edge detection operation through the edge detection algorithm of the Plaus operator to extract edge information of the part and generate an edge image;

[0059] S2-4: the preprocessing component converts the edge image into a binary edge image matrix B′;

[0060] Generate a binary edge image matrix B′:

[0061]

[0062] Among them, B′(x,y)=0 is the edge of the part;

[0063] B′(x,y)=1 is the background area;

[0064] S2-5: The preprocessing group inputs the edge image matrix B′ into the recognition component.

[0065] S3: the recognition component performs feature extraction and feature matching on the binary image data based on a preset image recognition model, and the recognition component obtains recognition data through feature matching;

[0066] Specifically: S3-1: The recognition component is provided with a contour detection model C, and the contour detection model C extracts key shape features of the edge image matrix B′ to obtain an extraction vector F:

[0067] Where F = C(B′)

[0068] F=[f1,f2,...,f n ]

[0069] f i is the part feature parameter;

[0070] S3-2: A standard part feature database is provided, the identification component is connected to the standard part feature database, and the standard part feature database has a standard vector group {S1, S2, ..., S m}, where S i is a preset standard feature vector in the standard vector group;

[0071] S3-3: The recognition component sequentially compares the extracted vector F with the standard vector group {S1, S2, ..., S m Each S in i By comparison, we can calculate F and each S in the vector group according to the following formula: i Similarity:

[0072] d(F,S i )=‖FS i ‖

[0073] S3-4: The recognition component obtains the best matching feature S c , according to the optimal matching feature S c , determine the state of the part and generate identification data R:

[0074] R = [id, sc, ds]

[0075] Among them, id is the identifier of the part category, sc is the similarity score, and ds is the defect type of the part.

[0076] S4: the identification component sends the identification data to the data analysis component;

[0077] S5: After receiving the identification data from the identification component, the data analysis component analyzes the identification data and generates an evaluation report for the automobile parts;

[0078] S6: The data analysis component sends the evaluation report to the control component;

[0079] S7: The control component automatically adjusts the working state of the detection system according to the evaluation report.

[0080] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0081] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. An automobile parts detection system based on image recognition, characterized in that: It includes image acquisition component, preprocessing component, recognition component, data analysis component and control component; The image acquisition component is used to acquire the image of the automobile part to be detected, and the image acquisition component communicates with the pre-processing component through a data channel; The preprocessing component is used to preprocess the received image and transmit it to the recognition component; The recognition component processes the received image data based on a preset recognition model and generates recognition data of automobile parts. The data analysis component is used to generate a quality assessment report of the automobile parts based on the recognition data. The control component is used to adjust the working state of the production line according to the quality assessment report.

2. The automobile parts detection system based on image recognition according to claim 1, characterized in that: The preprocessing component is used to perform grayscale conversion, denoising and edge detection on the received image.

3. The automobile parts detection system based on image recognition according to claim 2 is characterized in that: The identification data includes a part class identifier, a match score, and a defect status.

4. The working method of the automobile parts detection system based on image recognition according to claim 3 is characterized in that: S1: The image acquisition component acquires the image of the automobile part to be detected and transmits the image to the preprocessing component; S2: the preprocessing component preprocesses the received automobile part image and inputs the binary image data into the recognition component; S3: the recognition component performs feature extraction and feature matching on the binary image data based on a preset image recognition model, and the recognition component obtains recognition data through feature matching; S4: the identification component sends the identification data to the data analysis component; S5: After receiving the identification data from the identification component, the data analysis component analyzes the identification data and generates an evaluation report for the automobile parts; S6: The data analysis component sends the evaluation report to the control component; S7: The control component automatically adjusts the working state of the detection system according to the evaluation report.

5. The working method of the automobile parts detection system based on image recognition according to claim 4 is characterized in that: The S2 comprises the following steps: S2-1: The preprocessing component performs denoising on the grayscale image through the grayscale algorithm of the cv2.cvtColor function; S2-2: The preprocessing component eliminates noise in the image through a bilateral filtering denoising algorithm; S2-3: The preprocessing component performs edge detection operation through the edge detection algorithm of the Plaus operator to extract edge information of the part and generate an edge image; S2-4: the preprocessing component converts the edge image into a binary edge image matrix B′; Generate a binary edge image matrix B′: Among them, B′(x,y)=0 is the edge of the part; B′(x,y)=1 is the background area; S2-5: The preprocessing group inputs the edge image matrix B′ into the recognition component.

6. The working method of the automobile parts detection system based on image recognition according to claim 5 is characterized in that: S3-1: The recognition component is provided with a contour detection model C, which extracts key shape features of the edge image matrix B′ to obtain an extraction vector F: Where F = C(B′) F=[f1,f2,...,f n ] f i is the part feature parameter; S3-2: A standard part feature database is provided, the identification component is connected to the standard part feature database, and the standard part feature database has a standard vector group {S1, S2, ..., S m }, where S i is a preset standard feature vector in the standard vector group; S3-3: The recognition component sequentially compares the extracted vector F with the standard vector group {S1, S2, ..., S m Each S in i By comparison, we can calculate F and each S in the vector group according to the following formula: i Similarity: d(F,S i )=‖F-S i ‖ S3-4: The recognition component obtains the best matching feature S c , according to the optimal matching feature S c , determine the part status and generate identification data R: R = [id, sc, ds] Among them, id is the identifier of the part category, sc is the similarity score, and ds is the defect type of the part.

Images