复杂背景下改进的红外弱小目标检测

许多研究者关注红外弱小目标检测领域并进行过种种探索，然而复杂背景下检测的难题始终未得到满意的解决。复杂背景下的杂波难以消除，目标检测无法得到显著结果。为此，提出了一种基于高升压滤波器的加权三层窗口目标检测算法(high-boost weighted tri-layer local contrast measure, HB-WTLLCM),针对复杂背景的目标检测进行目标增强，从而提高检测率。本文算法首先利用改进的高升压滤波器对红外原始图像进行预处理，再利用三层嵌套窗口，根据目标形状进行局部对比度增强。最后引入一种基于复杂度评估的加权算法，进一步进行目标增强和随机噪声抑制。实验数据显示，本文提出的算法相比于主流算法在多建筑、多树木的复杂背景下目标增强能力更强，检测率更高。上述结果提示，本文提出的HB-WTLLCM算法对于复杂场景下红外弱小目标进行检测具有一定优势。

Research on Improved Infrared Dim Small Target Detection Under Complex Backgrounds

Recently, the field of infrared dim small target detection is paid considerable attention and various solutions are proposed. However, the problem of detection in complex background is unsolved. Clutter in complex backgrounds is difficult to eliminate, and significant results are difficult to obtain in target detection. For this reason, an improved algorithm, high-boost weighted tri-Layer local contrast measure (HB-WTLLCM) is proposed, which to enhance the target detection in complex background, and to improve the detection rate. An improved high boost filter was designed in this algorithm to preprocessed the original infrared image; then the tri-layer window was used for enhancing the local contrast. Finally, a weighted algorithm based on complexity evaluation was introduced for further target enhancement and random noise suppression. Experimental results show that, compared with mainstream algorithms, the proposed algorithm is stronger in target enhancement and improves detection rate under complex background of multiple buildings and trees. It is suggested that the HB-WTLLCM algorithm proposed in this paper can detect infrared dim and small targets well in complex background.