港口背景下红外运动目标的一种检测方法

为了解决港口背景下红外运动目标检测中受背景干扰带来的误分割和误跟踪问题,提出了一种基于港口背景抑制和光流检测的红外运动目标检测方法。首先，通过对小波分解图像进行OTSU分割，得到天水线区域。然后使用多级滤波确定序列图像中港口背景的抑制基准点，并根据这些背景抑制基准点实现序列图像的港口背景抑制。最后，运用光流预测实现红外运动目标检测。通过对实际港口背景红外图像进行背景抑制和红外运动目标检测的实验，验证了所提方法的可行性和有效性。

A Detection Method of Moving Infrared Targets against Harbor Background

To solve the problems of false segmentation and false tracking caused by the interference of background when moving infrared targets are being detected against harbor background, a moving infrared target detection method based on harbor background suppression and optical flow detection is proposed. First, OTSU segmentation is carried out for a wavelet segmented image, so as to obtain the sky-sea line. Secondly, a multi-level filter is used to locate the background suppression points in sequential images, so as to suppress the harbor background according to these background suppression points. Finally, an optical flow prediction technology is used to realize the detection of the moving infrared target. The feasibility and effectiveness of the method are verified in the experiment of background suppression and moving infrared target detection against the actual harbor background.