Research on Automatic Recognition Method of Artificial Ground Target Based on Improved HED

Automatic target recognition technology is an important research direction in the field of machine vision. Artificial ground targets, such as bridges, airports and houses, are mostly composed straight lines. The ratio geometric primitive lines to triangle area formed by their combination used feature quantity describe group lines, so characterize artificial target. In view shortcomings traditional edge detection methods, background suppression, non-prominent missing positions, etc., this paper proposed image method based on depth learning. By combining algorithm with improved HED network, real-time was completed. An automatic template matching proposed. This solved problem both homologous heterogeneous matching, which has theoretical value. First, were combined form primitives line group, then relationship determined using characteristic group. best found edge, homonymous points calculated. affine transformation matrix between two images obtained according points, accurate position found.