Object Detection in Image with Complex Background

Abstract

Abstr act. Object detection is the key technology in computer vision, with broad application prospects. Object detection has great research value and practical significance as a hot spot of video surveillance in recent years. This paper proposes an algorithm for ship detection in image with complex harbor background. We test the performance of several texture descriptors, and a region growing method based on contrast texture feature is proposed to implement sea-land separation. Then, we apply a method combined with adaptive threshold segmentation and shape analysis for offshore ship detection. Furthermore, the salient boundary template matching in the sea-land border area is used for docked ship detection. The experimental results show that our algorithm is able to implement ship object detection in complex image with good robustness and real-time performance.