Image Fusion Research Based On Compressive Sensing

Image fusion is to integrate image information from different sensors for a specific scene,in order to gain an access to the scene for a more reliable and comprehensive description.At present,image fusion has been widely used in medical imaging,remote sensing detection,security monitoring,target tracking,intelligent transportation and other fields.With the rapid development of senor technology,image fusion needs to process more and more data and traditional Nyquist sampling theorem cannot meet the demand of fusing lots of images in practical applications.Compressive sensing breaks through the limitation of Nyquist sampling theorem,which can realize the image signal sampling and compression at the same time.Applying the theory of compressive sensing into image fusion can greatly decrease the amount of image signal sampling data and reduce the storage and calculation cost of image fusion.This paper introduces the significance and development status of image fusion based on compressive sensing,and analyzes the basic theory of image fusion and compressive sensing in detail.What's more,this paper proposes two new algorithms of image fusion based on compressive sensing.(1)It presents a new block-compressed sensing image fusion algorithm based on lifting Wavelets.This algorithm proposes to use overlapped block method to block image for making an improvement of the traditional image block method,and designs a new regional energy integration rule.This paper integrates three groups of different types of images and tests the performance of algorithms at different sample rates.By comparing the experimental results,it proves this algorithm in the paper can improve the image quality effectively.(2)It proposes an algorithm of image fusion based on compressed sensing of Nonsubsampled Contourlet Transform(NSCT)and Discrete Wavelet Transform(DWT).This algorithm takes use of the advantage of DWT which can extract the image details effectively and makes up the disadvantage of lacking of detail information of NSCT.Experimental results show that this algorithm makes the integrated image quality become better in the subjective and objective aspects,and the final images have more detail features and clearer profiles.Finally,the paper makes a summary and outlook.