关键词: 机器视觉, 烟标数量检测, Canny边缘检测算子, Hough变换

Abstract: There is an emerging requirement for quality control like sheet-counting in printing and packaging industry, where machine vision methods are widely adopted due to their nondestructive and real-time merits. However, the conventional single-camera scanning systems often suffer from insufficient resolution while handling the sheet-counting issue of ultrahigh stacks. To solve the problem, this paper presents a new apparatus using camera array and advanced image processing algorithms. The system architecture of the apparatus is briefly described, and much emphasis is exerted on the line detection of images based on an improved Canny edge algorithm. Particularly, the fusion of results from multi-camera imaging is described in details. Through more than one year’s test in factory production lines, it is demonstrated that the apparatus can satisfy the high-accuracy requirement of real-time sheet-counting task with an error less than 6.

Key words: machine vision, packing paper counting, Canny detector, Hough transform