Study On Compressed Sensing Image Fusion Optimization Algorithm Based On NSCT Transform

Image fusion refers to the process which makes multiple source images integration into a new image with more comprehensive information by appropriate technical means. It takes full advantage of complementary information between multiple images, removes redundant information and improves the accuracy and reliability of the image information. It provides a strong guarantee for people in related fields getting down to image processing, analysis and diagnostics.In conventional image processing, we often require high-resolution imaging devices to take pictures of the actual scene and then get massive image information. Especially, the amount of collected image signal in the field of image fusion, such as satellite remote sensing and geographical exploration, is pre-acquired enormous, and sometimes actually reaches intolerable degree. However, in practical applications, since the information between the images has great redundancy, most of the image information is discarded by compression coding. Therefore, to capture images from the source point of view, a lot of image information acquired by the imaging device has not been fully utilized eventually, but in vain wasted.Based on the above, the proposed compressed sensing provides a new way for people to study how to reduce the capture image information and avoid unnecessary waste of information resources. The theory makes the sensor capture compressed image information, direct acquire the sparse representation of the image, thereby avoiding a large number of useless information collected, saved into the sensor device. Meanwhile, it also reduces the pressure of storing and transferring huge amount of information in the subsequent process. Thus the main work of this paper is to apply compressed sensing into traditional image fusion.Firstly, it briefly describes the basic theory of compressed sensing, including the sparse representation theory, measurement matrix selection and recovery algorithm, makes several typical recovery algorithm for the simulation of two-dimensional image contrast. Secondly, it makes the analysis of the traditional image fusion theory, including the concept of image fusion, hierarchical, rules and integration evaluation, focusing on the traditional fusion algori- thm, from the early fusion algorithm based on spatial domain, again to decomposition based on tower, again to fusion algorithm based on wavelet transform and multi-scale transformation. Finally, it makes a detailed description of the image fusion method based on non-sampling contourlet transform (NSCT) of multi-scale transformation. Combined compressive sensing theory, it acquires compressed sensing image fusion optimization algorithm based on NSCT transform. And then make simulation contrast with traditional types of image fusion algorithm and proposed several fusion algorithm based on compressed sensing, reducing the amount of the original image data, acquiring a good fusion effect.