Mixed Linear Model Of Multi-channel Remote Sensing Image Pixel Decomposition Method

Usually, ground objects in remote sensing images are detected by units of the pixels. Due to the limit of spatial resolution, in most cases, one pixel may cover hundreds of square meters with various ground objects and becomes a mixed pixel. The mixed pixel problem not only influences the precision of object recognition and classification, but also becomes an obstacle to quantification analysis of remote sensing images. This problem can be overcome by precisely obtaining the percentages of object of interest. In fact, the exact decomposition of mixed pixels is very important in the field of subpixel classification of multispectral/hyperspectral remote sensing image as well as detection and identification of ground objects. Focusing on the aforesaid issue, the article has made a lot of research, and the innovations are as follows:1. A new algorithm for decomposition of mixed pixels based on orthogonal bases of data space is proposed in this paper. It is a simplex-based method which extracts endmembers sequentially using computations of largest simplex volumes. At each searching step of this extraction algorithm, searching for the simplex with the largest volume is equivalent to searching for a new orthogonal basis which has the largest norm. The new endmember corresponds to the new basis with the largest norm. This algorithm runs very fast and can also guarantee the consistence of its final results. Moreover, with this set of orthogonal bases, the proposed algorithm can also determine the proper number of endmembers and finish the unmixing of the original images which the traditional simplex-based algorithms cannot realize by themselves.2. Mixed pixel decomposition in highly mixed data is a very difficult problem. This paper presents a new scheme based on Nonnegative Matrix Factorization (NMF) and simplex-based method to solve this problem. In addition, some appropriate constrains are introduced into NMF for the decomposition of mixed pixels. Experimental results obtained from both artificial simulated and real-world remote sensing data demonstrate that the proposed scheme for decomposition of mixed pixels has excellent analytical performance.3. This paper proposes a method for fully constrained abundance estimation based on noise estimation. This method demonstrates the geometric explanations for constrained least squares estimation. By introducing the noise estimation process and condition judgment mechanism, the proposed method can avoid over-fitting effectively and has a better computing efficiency.