Abstract: In order to determine the abrasive wear on cigarette packing film, a method was proposed based on machine vision technology. A friction coefficient meter was modified to simulate the abrasive wear on the film, and a digital camera was used to collect the image of worn film. All the characteristic data relevant to the film image were obtained through edge contour extracting, binary processing, image segmentation and denoise processing, and the degree of abrasive wear was calculated by an image processing software. The results indicated that this method was accurate, stable and reliable for determining the abrasive wear of cigarette packing film with the measurement error of 0.1%.