提出了一种基于多尺度形态学和连接相对熵的红外图像自动目标检测方法;它首先对目标红外图像进行多尺度形态滤波,消除背景和杂波的影响同时增强目标的对比度,然后采用连接相对熵准则确定区分目标和背景的最佳门限,最后对得到的二值图再进行形态滤波,得到精确的目标分割图像。实验结果表明该方法具有一定的鲁棒性和自适应性,能够实现目标的自动、非参数化的有效检测,且效果良好,便于下一步的目标识别和跟踪。 An infrared image automatic target detection method based on multi-scale morphological and connection relative entropy is proposed. Firstly, multi-scale morphological filtering is performed on the target infrared image to eliminate the influence of background and clutter while enhancing the contrast of the target. Then, The entropy criterion determines the optimal threshold for distinguishing the target from the background. Finally, the morphological filtering is performed on the obtained binary image to obtain an accurate target segmentation image. The experimental results show that the proposed method has some robustness and adaptability, and can achieve the automatic and non-parametric effective detection of the target with good effect, which is convenient for the next target recognition and tracking.