Ceramic Ring Pieces Gap Detection And Size Measurement System Based On Machine Vision

The fast developing of computer hardware and data processing capacity enables image processing application booming in industry area. Machine vision is designed to replace humans with industrial cameras and algorithms for industrial measurement and detection.This thesis reported the work of ceramic ring pieces gap detection and size measurement. Two30-megapixel industrial cameras were used to detect notches of the inner periphery and the outer periphery of the ceramic ring. The thickness of the ceramic ring was also accurately measured using camera calibration. The system runs in real-time with high reliability.The target object was first separated from background using color information. The RANSAC algorithm was then used to fit circles and to find circle centers from edge information. Gaps were detected by central symmetry. The detection accuracy is98%for parts with big gaps.For measuring the height of a part, we used the calibrated cameras. By searching the bottom point of the cylindrical part and the corner point, the vertical height of the part can be accurately measured. The measurement accuracy reaches500millimeter.