A maximum likelihood algorithm for detecting line segments
Qi Sun, Minghao Yang · 2010
The detection of lines in an image is an important task. A maximum likelihood algorithm for detecting line segments and curve segments is presented in this paper. The theory of the proposed method is that adjacent pixels are connected into segments when their tangent directions are nearly equal. Tangent direction of edge pixels is estimated by enumerating predetermined masks on several directions. In addition, the acceleration strategy is given, which makes computing cost much closer to enumerating a spatial gradient operator. Experience results show that the method can well determine the disconnected points in lines. Furthermore, it is more effective in terms of input dependence and time cost for detecting curve segments in contrast edge than traditional methods.