Research On Detection System Of Cigarette Case Based On Machine Vision

The Detection System of Cigarette Case is a special device, which is used to inspect the packaging quality on the cigarette product line. It is a complicated system relating to photo-electricity, optics, computer, electronics and mechanism, etc. The apalication of the system greatly enhances the level of automation and lessens the stress of labor. In addition, it also ensures the quality of final products. This system can be widely applied in tobacco industry, pharmacy industry, foodstuff industry, etc.This paper mainly researches on the Detection System of Cigarete Case based on machine vision. Researching on digital image processing technology and considering the locale condition, the paper gives the detail arithmetics for Templates-Matching, Line-Detection, Side-Detection and Pattern-Detection. Kernel-Matching based on region-of-interest is used to reduce the matching time compared with the traditional method. The paper gives an effective arithmetic based on least-squares to inspect the Line. This approach fits a line equation by calculating the max grads, and judges the Line by the parameters. To get the suitable images, the paper makes researches on the illumination system about lamp-house, lamp way, illumination technique, and the grabbing system about CCD, lens, image-acquire-card, photoelectric switch. The technology of VPN is also introduced for remote control and remote upgrade. And this makes the remote diagnosis and remote maintenance possible.