论文部分内容阅读
针对复杂背景红外序列中小目标检测问题,提出了基于形态学Top-hat算子和运动连续性的目标检测算法。形态学中的Top-hat算子是一种极好的高通滤波算子,开Top-hat算子能检测图像中的峰,而闭Top-hat算子则能检测图像中的谷。通过对Top-hat算子选择合适的结构元,就可以将候选的目标从复杂的背景中提取出来,然后利用序列图像中运动目标的连续性和轨迹的一致性筛选出真正的目标。试验结果表明,该方法能有效抑制背景噪声、检测并跟踪红外小目标。
In order to detect small targets in complex background infrared sequences, a target detection algorithm based on morphology Top-hat operator and motion continuity is proposed. Morphology Top-hat operator is an excellent high-pass filter operator, Top-hat operator can detect the peak in the image, and closed Top-hat operator can detect the image of the valley. By selecting suitable structure elements for Top-hat operator, candidate objects can be extracted from complex background, and then the real targets can be screened by the continuity of moving objects and the consistency of trajectories in sequence images. Experimental results show that this method can effectively suppress background noise, detect and track infrared small targets.