Research on real-time error measurement in curve grinding process based on machine vision

A machine vision image measurement system for online monitoring of the wheel wear degree during the curve grinding process is designed and developed. The measurement apparatus and its principle of operation are introduced in detail. Real-time image of work piece and wheel in the grinding zone is gathered by CCD camera installed in the grinder. For the purpose of increasing the measurement precision, a new edge detection approach combining Zernike moments operator with Prewitt operator is proposed. The edge of the finished work piece is located with sub-pixel level accuracy, and then the machining error of the work piece is calculated on-line by comparing with the theoretical curve of the work piece. An application of its validity and the experimental results are also given. Experimental results demonstrate the proposed measurement method in this paper is effective, and its detection precision and results are reasonable.