Recent studies indicate sparse representation has achieved very good results in hyperspectral imaging tar-get detections. However, sparse representation only uses the spectral information of the hyperspectral imaging, with-out considering the spatial information. The purpose of this paper is to research the influence of hyperspectral ima-ging's spatial correlation on the target detection algorithm, and propose two new sparse representation models, which named 4-neighborhood smoothing sparse model and 4-neighborhood union sparse model, considering sparse representation of both target pixels and its 4-neighborhood pixels to improve the effectiveness and efficiency of the target detection algorithm. With three real hyperspectral imaging datasets in the simulation experiment, the results show that the algorithms utilizing the new models could improve the effectiveness or efficiency of the test results to a certain degree.%稀疏表示方法在高光谱图像目标检测中取得了较好的检测效果，但其只利用了图像的光谱信息，没有考虑空间信息。针对高光谱图像的空间相关性对目标检测算法的影响，提出了分别采用4-邻域平滑稀疏模型和4-邻域联合稀疏模型对高光谱图像进行目标检测的算法，将目标像素及其4-邻域像素的稀疏表示综合考虑，提高检测算法的效果和效率。使用3组高光谱图像数据进行了仿真实验，实验结果表明，所提出的2种方法分别在检测效果和计算效率上有一定程度的提高。
展开▼
