Counting Method for Identifying Vessels at Sea Based on Contour Extraction Recognition

In this paper, we propose a counting system based on contour extraction and recognition for identifying ships at sea, by controlling the range of pixel values of the gray scale image in order to weaken the influence of environmental factors such as clouds and waves on the image. And the Canny algorithm is modified by testing a large number of images so as to extract the contour of the vessel from the sea surface image. The contour lines are processed by drawing frames after corrosion and expansion, and the designed feature extraction and feature matching algorithms are used to identify and count the ships, so as to realize the identification and counting of ships at sea from the visual point of view. Experiments prove that the method has good detection effect on the water environment under different weather.