Feasibility Study of Compressive Sensing Underwater Imaging Lidar.

摘要

This final technical report describes the efforts during the interval from July 2012 through the end of December 2013 to accomplish the project objective of studying the feasibility of developing a compressive sensing theory based approach to form a grayscale image in the underwater environment. This objective was accomplished through a series of theoretical work, simulations and experimental studies. In the first stage of this project, the frame based Compressive Sensing underwater imaging system concept was further studied. This study resulted in a publication that was among the top most downloaded papers on the Journal of Electronic Imaging website until November 2013. One of the most important achievements of the project is that the Compressive Line Sensing underwater imaging system was conceptualized. Built upon the theoretical foundation the Compressive Sensing theory and the Distributed Compressive Sensing theory, this technique employs the paradigm of independently sensing each line and jointly reconstructing a group of line. Such implementation is compatible with the traditional whiskbroom type survey, but achieves significant resource compression by exploiting the correlation/redundancy among adjacent target regions. An image reconstruction framework was developed based on this concept. Extensive simulations were conducted to validate this framework. A prototype system was developed to further validate the Compressive Line Sensing framework through the experimental study. A journal paper describing this effort will be published in Journal of Optical Engineering special section on ocean optics.