Study Of Target Detection In Remote Sensing Imagery

This paper deals with target detection in remote sensing imagery, which involves the affection of atmosphere to remote sense, target detection in single channel and target detection in hyperspectral respectively. Firstly, an atmospheric correction algorithm based on 6S model is developed by considering the position relationship between the sensor and the earth surface and the influence of atmosphere absorption and scattering. Secondly, a texture-based target detection approach is presented to overcome the limitation of the cluster-based approach. The new texture-based approach segments one image into different textures and analyzes the distribution of pixel values of the textures. Thirdly, the two dimensional texture segmentation model is extended to three dimensional texture segmentation to describes the background property of hyperspectral imagery and a new hypespectral target detection algorithm is proposed by introducing the conventional RX detector to complement the segmentation result. Extensive experiments applied to real images of small target and extend target validate the good performance of the approach.