一种利用均值匹配改进的高光谱异常检测方法

传统高光谱异常检测算法由于背景信息估计不准确等原因普遍存在高虚警率的问题,针对这一现象,提出了一种利用图像均值进行匹配改进的高光谱异常目标检测后验处理方法。首先采用传统的高光谱异常检测算法将待检测高光谱图像划分为背景与异常目标潜在区域,之后通过对待测图像求解均值,将其与异常目标潜在区域像元进行相似性匹配计算,剔除大范围误检像元,得到最终检测结果。该方法在传统异常目标检测算法基础上进行相似度量剔除大范围虚警像元,在提高原算法探测能力的同时有效地降低虚警率。实验表明,该方法可以有效降低虚警率,提高原算法对于亚像元异常目标的检测能力,且对于不同算法、不同数据具有普适性。

Improved Hyperspectral Anomaly Target Detection Method Based on Mean Value Adjustment

High false alarm rate issue generally exists in traditional hyperspectral anomaly target detection algorithms due to inaccurate background information and other problems.An improved hyperspectral anomaly target detection method based on mean value adjustment is proposed to solve the high false alarm rate problem in traditional hyperspectral anomaly detection algorithm.Firstly,the traditional anomaly detection algorithm is used to divide the detected image into the background and the potential area of the abnormal target.Then we calculate the mean value of the whole image and the similarity of pixels in potential area are compared with it.Finally,the false pixels are removed according the similarity of pixels in potential area and mean value.The proposed method improves the detection ability of traditional algorithms and reduces the false alarm rate of traditional algorithms.The experimental results indicate that the proposed method can reduce false alarm rate and improve the ability of detecting sub-pixel anomaly targets than traditional algorithms effectively.It also has robustness for different algorithms and different data.