An Algorithm for Detecting Lines Based on Primitive Connection

Abstract

Line detection in gray-scale images is a low-level processing which has many applications such as road detection in remote images and vessel detection in medical images. An efficient primitive connection-based algorithm for detecting line segments is proposed. Canny operator is used to detect edges in a digital image and all the edges are tracked to form series of edge point chains by a 4-neighbor-priority eight-direction tracking method. These edge point chains split in shorter ones to produce primitives. In the end, line segments are produced by merging several primitives, which satisfy several qualifications. Both synthetic and real images have been tested by the proposed algorithm and the experimental results show that it is efficient.