Application Of Abnormal Spectral Image Abnormality Target Detection And Spectral Imaging In Camouflage Evaluation

The high spectral resolution of hyperspectral image makes it have good performance in the spectrum diagnosis, so it is very suitable for the extraction of artificial target from natural background. Among them, the hyperspectral image anomaly detection algorithm does not need a priori spectral information; it can detect the target in spectral differences with background, has stronger practicability, and gradually become a hot research topic in the field of target detection. At the same time, the rise of hyperspectral imaging reconnaissance technology has brought the huge challenge on traditional military camouflage technology, so the spectral imaging technique is introduced into the evaluation of camouflage effect, make with diversified war, can be more comprehensive survival ability and targets on the battlefield covert performance assessment, has the extremely important practical significance. This paper focuses on hyperspectral anomaly detection algorithm and application research of the spectral imaging technique in the evaluation of camouflage effect based on the in-depth analysis under the structure of data of hyperspectral image.First of all, by analyzing the traditional anomaly detection algorithm of hyperspectral image, to find out its insufficiency, in view of the shortcomings, the hyperspectral image anomaly detection algorithm based on dimension reduction algorithm is studied. And study of hyperspectral image data dimension reduction algorithm, the dimension reduction algorithm based on minimum noise fraction transform is improved. The most important step of minimum noise fraction transformation is to estimate the noise covariance matrix, in this paper, the noise covariance matrix is estimated by neighborhood weighted mean method and Canny operator, obtain good effect of dimension reduction and anomaly detection results.Secondly, the spectral imaging technique is introduced into the camouflage effect evaluation. Put forward a kind of camouflage evaluation method based on spectral information, and define the Euclidean camouflage coefficient for the quantitative analysis of camouflage, verify the camouflage evaluation method through the anomaly detection algorithms, the results show that this method can effectively reflect the goal against spectral imaging reconnaissance camouflage effect. At the same time, the Euclidean camouflage coefficient calculation software is designed.