Image analysis-based online detection method and system for injection mold of automobile parts

Abstract

The application discloses an image analysis-based online detection method and system for automobile part injection molds, and relates to the technical field of automobile part injection mold defect detection, and comprises the following steps: acquiring an image of an automobile part injection mold to be detected, and marking the image as a detection mold image; performing gray image conversion on the detection mold image to obtain a detection gray image, performing binaryzation processing on the detection gray image to obtain a detection binaryzation image, and performing contour extraction on the detection binaryzation image to obtain a defect contour; acquiring the RGB threshold value of all pixel points based on a normal automobile part injection mold image; acquiring the position of the defect contour on the original detection mold image based on the position of the defect contour in the detection binaryzation image; and the application is used for solving the problem that in the existing automobile part injection mold defect detection technology, when a mold image without defects is acquired, the change of the RGB value of a pixel point under different illumination conditions is not considered, and the automobile part injection mold defect detection is not accurate.

Images