No abstract

No abstract

No abstract

In view of the problem that the measurement accuracy of screen gaps is affected by the R region of the transition rounded corner of the large-size flat back cover, a mobile vision system with dual linear motor synchronous drive is designed to realize precise positioning of monocular vision in 16 local image acquisition positions set in the R region. Full-size measurement of the R region is achieved using local information fusion to accurately measure the screen gap. First, to solve the problem of edge breaking and the blurring of local images, a region extraction and segmentation method based on block statistics is proposed to quickly obtain accurate outer edges. Then, an adaptive Hough line detection method based on dichotomy is proposed to accurately locate the outer edges. The experimental results show that the accuracy of region extraction and segmentation is up to 99.68% within 60 ms; the angle error and position error of line detection are less than 0.01° and 0.2 pixels, respectively. With optimal anti-interference performance, the proposed algorithm realizes the adaptive and accurate positioning of the outer edge within 200 ms; the measurement time and the measurement error of the measurement system are less than 6.0 s and 0.03 mm, respectively.