Die casting surface defect detection method and system based on machine vision

Abstract

The invention provides a die casting surface defect detection method and a die casting surface defect detection system based on machine vision. The die casting surface defect detection system comprises an image acquisition unit, an image processing unit, a HOG + LBP feature extraction unit and an SVM classifier training detection unit, the image acquisition unit is used for acquiring images of thesurfaces of die castings; the image processing unit is used for processing the collected images to obtain images capable of extracting defect feature vectors; the HOG + LBP feature extraction unit isused for performing HOG + LBP feature extraction on the processed image; and the SVM classifier training detection unit is used for training the SVM classifier and carrying out defect identificationby using the trained SVM classifier. According to the method, HOG + LBP feature calculation is adopted, complex multi-dimensional feature calculation is effectively avoided, recognition efficiency isreduced, recognition and classification are carried out through an SVM classifier, the number of needed samples is small, and the problem that a neural network algorithm falls into a local minimum value when sample types are complex is solved.

Description

technical field [0001] The invention relates to the technical field of automatic detection of machine vision defects, in particular to a method and system for detecting surface defects of die-casting parts based on machine vision. Background technique [0002] Die casting is a casting method in which liquid metal is filled into a die-casting cavity at high speed under high pressure and solidified under pressure to form a casting. The die-casting production process is relatively complicated, and many defects will inevitably occur on the surface of the casting. At present, customers' requirements for product quality are increasing, and higher requirements are put forward for the quality inspection and quality information collection of die-casting products. [0003] The detection of surface defects of aluminum alloy die castings by major domestic die-casting factories mostly relies on manual detection, the efficiency of manual detection is relatively low, the workload is large ...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the problem of low efficiency of manual detection and X-ray detection, and the inability of the visual detection method in the prior art to be effectively applied to the surface detection of die-casting parts, the present invention proposes a method and system for detecting surface defects of die-casting parts based on machine vision

Images