Removing Natural Objects from the Sea Surface Background Image Based on Contour Map and Local-Hausdorff Distance

Abstract

Objective: Using the target classification is too complex be used to distinguish between natural objects and Man-made objects in aerial image, so in this paper, we analyze the target classification and feature of the target in the sea instead. According to the features of the contour map, a method for eliminating natural objects is proposed. Methods: First, we decompose the contour map through perspective transformation, to establish the estimate of the true image, second, the Fourier-mellin transformation is adopted to match the outline of estimate object image with the outline of true object image (matching), and then we use Hausdorff distance to correct the preliminary matching, Finally, we adopt the difference method to remove the natural objects. Results: The experimental results show that the method can effectively remove the natural objects and keep the Man-made objects in the observed image. Conclusion: In this study, we propose a new method of removing the natural objects based on the known contour information of perspective transformation and It has been proved by the experimental results that this method is feasible for target detection by reducing the difficulty of target detection and improving the efficiency of target classification.