On-line detection system and detection method for surface defects of parts based on feature matching

Abstract

The invention discloses an on-line detection system and detection method for surface defects of parts based on feature matching. The system includes an image acquisition device and a computer. The image acquisition device includes a CCD camera and a light source connected to the computer. The computer includes a software client, an image processing module, and a detection module. The software client is used for realizing the control, detection and display functions of online detection of the surface defects of parts and is connected to the image processing module. The detection module is connected to the image processing module. According to the system, a Hu moment value is calculated based on the Sobel characteristic map of the part and similarity matching is performed to complete surfacedefect detection of the part. The image acquisition device shoots a part image; the image processing module preprocesses the image, carries out segmentation, and extracts features; the detection module calculates the similarity by the Hu moment value and outputs a determination result. The system has a simple structure and is suitable for the detection scene of part defects in the field of industrial inspection; and real-time online detection of the high-precision part surface defect is realized.

Description

technical field [0001] The invention relates to the fields of industrial inspection and computer vision, in particular to an online detection system and detection method for surface defects of parts based on feature matching, which is applicable to the detection scene of parts defects in the field of industrial inspection, and can perform high-precision online detection of surface defects of parts in real time . Background technique [0002] There are many kinds of surface defects of parts, such as flash, lack of material, scratches, stains, etc., and these defects need to be detected. The inspection process in traditional parts production is generally carried out by manual inspection, which not only reduces the production efficiency of the assembly line, increases the labor burden, but also fails to detect defects quickly and accurately. Therefore, it is a problem that needs to be solved in this field to detect surface defects of parts by computer vision system to replace ...

AI Technical Summary

Problems solved by technology

Grayscale information does not necessarily reflect defect information, but may also be information such as noise, brightness, and color difference. It is impossible to determine whether the difference in the histogram is caused by defects in statistical features

[0006] 3. The biggest problem with the detection method based on defect features is that it cannot cover all defect types. Even if all known defect features are already in the library, new defect types may appear during the detection process; moreover, even if the same defect The features also have different feature representations, and it is impossible to cover all situations

Images