A suppression method of image noise for cylindrical objects based on vision sensors

Abstract Measurement based on vision sensors plays a significant role in process control and product quality in manufacture process. In extreme conditions, the ability to effectively distinguish information from noise is essential in the measurement process. In this paper, an online detection method that achieves noise suppression based on credibility evaluation of centre points is proposed which is applicable to cylindrical objects. A correlation expression between energy and relative position of the cross-section of the light stripes is also derived, and a method for determining the credibility of centre points with the greyscale distribution rule of the correlation expression is established. Experimental results indicate that the greyscale distribution rule for light stripes is reliable, and the noise suppression method can eliminate the noise of cylindrical objects. In the experiments conducted, the relative error of laboratory measurements using the proposed method reduced from 1.84% to 0.33%.