Line Feature Recognition Based on Differential Projection Transform

A new method is introduced in line feature recognition, firstly introduced the theory and algorithm of differential projection transform and its inverse transform, and then compared and analyzed the differential projection transform, Radon transform and Hough transform in aspect of accuracy of line feature recognition and asymptotic time complexity, and draw a conclusion that the differential projection transform is better than Radon transform and Hough transform in thick line recognition, and gives the general steps in line feature recognition using differential projection transform.