Cigarette Filter Rod Parameter Measure Based on Otsu Image Processing

Abstract

Traditional contacting measure method can easily cause cigarette filter rod distortion and needs long measure time. Its measure precision is difficult to satisfy practice demand and it cannot implement online measure. In order to overcome these shortages, imaging measure technology is used to measure cigarette filter rod parameters. Firstly, preprocessing filtering is operated on the image acquired from imaging measure system. Then binary cigarette filter rod target image is obtained through Otsu segmentation method. After that target feature extraction is processed and correlative parameters of cigarette filter rod are gained. Parameters measuring absolute error for two images which are acquired from the same cigarette filter rod under different illumination is 0.683%. Experiment proves that the proposed method can rapidly and correctly measure cigarette filter rod and has good robust and adaptability. It can effectively measure cigarette filter rod in practice.