ABSTRACT
At present, the domestic cigarette manufacturers pay more attention to the quality of the production of cigarettes. Quality of cigarette rod can be tested e- ciently and real-time by using machine vision technology. Roundness error detection, largely reflects the quality of cigarette production. (Huang 2001, Rosin 1999)
Machine vision technology, which combines optics, computer science and technology and digital image processing technology is an online non-contact measurement method. The principle of main machine vision measuring roundness error is through CCD (Charge Coupled Device) camera to take images of the measured object, and use digital image processing technology for processing, then extract the image features, analysis and process it, get the rounding error of measured object (Liu & He 2007), further evaluate the roundness error of the measured object. Walsh detected the appearance, quality of a paper cup with machine vision technology, and achieved the paper cup online non-contact measurement and classification (Walsh & Raftery 2012). Sharma used the machine vision technology to measure inner diameter roundness of the optical fiber, realized fast and non-contact measurement of optical fiber diameter roundness (Sharma 2010). This method has advantages of non-contact, high efficiency, high measuring precision, it becomes the main online measuring method in the non-contact measurement field.
