Workpiece defect rapid detection method based on contour features

Abstract

The invention relates to the field of hardware workpieces, and discloses a workpiece defect rapid detection method based on contour features. The method includes the steps that firstly, information of a template workpiece image is pre-stored, and two contours are selected in the template workpiece image; secondly, position information of each contour of the detected workpiece image is obtained by using a cfFindContours function; thirdly, the information of the detected workpiece image is read, and two contours same as the contours in the template workpiece image are selected, center points corresponding to the two selected contours are found out, and the slope of a connection line of the center points of the two selected contours is worked out; fourthly, the slope serves as the rotating reference of the to-be-detected workpiece image, so that the detected workpiece image is aligned with the template workpiece image; fifthly, the detected workpiece image is matched with the template workpiece image through an image difference method, and whether a workpiece has defects or not is detected. By adopting the scheme, the defects on the detected workpiece can be rapidly found, and the detection speed is effectively increased.

Description

Technical field [0001] The invention relates to the field of hardware workpieces, in particular to a method for quickly detecting workpiece defects based on contour features. Background technique [0002] In industrial production, due to the movement of the conveyor belt, it is inevitable that there will be a position error between the detected workpiece and the template workpiece. In order to improve the accuracy of detection, it is necessary to adjust the position of the detected workpiece to align the position of the detected workpiece with the template workpiece, and then use the image difference method to perform image matching to identify the passing and failing workpieces. [0003] One method of image alignment is to find the line between the two contour centers in the workpiece, and use the slope of this line as the standard to rotate the inspected workpiece and the template workpiece so that the inspected workpiece and the template workpiece are at the same angle. Then pu...

AI Technical Summary

Problems solved by technology

Contour centroid calculation formula: Among them, I(i, j) is the set of grayscale image pixels. In the actual detection process, due to the phenomenon of reflection, shadow and contour edge connection on the surface of the hardware workpiece, it will directly affect the centroid position of the contour calculated by the pixel points. Accuracy, resulting in errors in the alignment between the template and the detected image

On the other hand, when the image resolution is very high, the data volume of the contour in the image will increase, so that the calculation of the contour centroid will increase, and the detection speed will decrease.

Images