Hough Transformed Based Ship Segmentation Using Centerline Extraction and Feature Angles

Image segmentation is a very important step in order to accomplish automatic tasks, such as classification, being continuously demanded by our society. However, the particular case of naval images entails issues that may exacerbate the conditions for segmentation. Specifically due to water reflections the gray distribution might be very disperse throughout the image and waves produce a large amount of edges. They differ from one image to another too and thus, patterns of color or brightness are difficult to be found. Nevertheless, in this paper we take advantage of these negative characteristics by applying a Hough Transform based technique that enhances the shape of the ship. First, the ship centerline is extracted and the desired feature angles are applied based on this value. The preliminary results are further processed to obtain higher accuracy. The set of parameters is presented and discussed, as well as the non-maximal suppression step and the voting strategy employed. The results reach a higher level of detail compared to previous suggested techniques for this kind of images.