Hyperspectral Image Target Detection Algorithm Based on Complex Networks

The broad application of hyperspectral devices makes hyperspectral image target detection a popular research. This paper presents a new method for target detection by transforming the hyperspectral data into complex networks. By analyzing the networks topological feature corresponding to each pixel, one can easily evaluate the statistical characteristics and intrinsic properties of the signals. Thereby matching the target by comparing the pixel network characteristics. The spectral characteristics of the pixel are analyzed from the perspective of the network, which provides a new spectral matching criterion. Experimental results demonstrate that the proposed method can acquire satisfactory results when compared with traditional methods.