Generalized hough transform for object classification in the maritime domain

摘要

A generalized Hough transform-based classification scheme for an object-of-interest in maritime-domain images is proposed in this paper. The scheme explores the use of Hough features and neural networks to classify large sets of image objects collected in the maritime domain environment. The object edge points are extracted and used to generate the generalized Hough coordinate tables. The Hough coordinates are in turn reformatted to form Hough features maps. The coordinates of dominant peaks called Hough features are extracted and fed into a feed-forward, back-propagation neural network for classification. In this research, the scheme is tested using perfect geometric shapes as well as maritime-domain images of ships, aircraft, and clouds, and the classification results obtained are reported.