Fast infrared maritime target detection: Binarization via histogram curve transformation

Abstract To improve the accuracy and efficiency of infrared maritime target detection under different environmental conditions and for different kinds of targets, we proposed a novel self-adaptive binarization algorithm which is based on the histogram curve transformation. The main contribution was a rapid and robust method for detecting infrared maritime targets that have positive local contrasts. This method has low computational complexity and high detection accuracy under a variety of conditions and enhances the accuracy and speed of single-frame detection for infrared maritime distressed targets. The proposed histogram rightwards cyclic shift binarization (HRCSB) first transforms the histogram curve (HC) according to a self-adaptive gray level transformation equation. Then, the background subtraction based on Gaussian filtering can be used to generate an enhanced image. Finally, the final HC can be extracted from this enhanced image. After a cyclic shift of the final HC, the average gray level of the shifted HC can reveal an effective threshold for detecting targets from the enhanced image. Experimental results show that, compared with four existing algorithms, the proposed HRCSB can successfully detect targets under a variety of conditions while keeping a low false alarm rate and a low computational complexity. Thus, the proposed HRCSB algorithm has potential for excellent applicability.