论文部分内容阅读
针对伪装目标检测问题,提出了一种有监督的高光谱伪装目标检测方法。以植被型伪装目标为研究对象,在分析伪装材料与绿色植被光谱之间特性的基础上,先通过光谱重排、光谱微分以及光谱差异性增强处理,对植被型伪装材料与真实植被(背景)之间的光谱差异进行放大,然后利用主成分分析(PCA)变换进行降维,从而实现了一种适用于大面积植被型伪装目标的高光谱检测方法。实验结果表明,该检测方法在检测时间和检测效果上要优于基于加权的约束能量最小化法(WCM-CEM)和基于非监督目标生成处理的正交子空间投影法(UTGP-OSP)。
In order to detect camouflage targets, a supervised hyperspectral camouflage target detection method is proposed. Taking the vegetation-type camouflage target as the research object, based on the analysis of the characteristics between the camouflage material and the green vegetation spectrum, the vegetation-type camouflage material and the real vegetation (background) are firstly analyzed by spectral rearrangement, spectral differentiation and spectral diversity enhancement, , Then the principal component analysis (PCA) transform is used to reduce the dimension, so that a hyperspectral detection method suitable for large-area vegetation-type camouflage targets is achieved. The experimental results show that this method is superior to WCM-CEM and UTGP-OSP based on unsupervised object generation in detection time and detection.