A Semi-Supervised Rank Tracking Algorithm For On-Line Unmixing Of Hyperspectral Images

This paper addresses the problem of rank tracking in real time hyperspectral image unmixing methods. Based on the On-line Alternating Direction Method of Multipliers (ADMM), we propose a new hyperspectral unmixing approach that integrates prior information as well as joint sparsity regularization, allowing to select only the active components on each sample of the image. This results in a semi-supervised algorithm, well adapted for real time rank tracking for on-line acquisition systems (such as the pushbroom imager). Experimental results on synthetic and real data sets demonstrate the effectiveness of our method for parameters estimation and rank change detection.