Research Of The Pathological Medical Image Compression Algorithm Based On Compressed Sensing

With the rapid development of science and technology,super-resolution imaging and other system has been widely applied to medical pathological tissue image diagnosis,and the telemedicine system has also received extensive attention and application,which makes the disease easy to diagnose,but makes the transmission and storage of pathological medical images become an urgent problem to be solved.The development of traditional image compression and transmission is limited by the Nyquist Sampling theorem,so that the problem cannot be solved effectively,but the Compressed Sensing(CS)is not bound by the Nyquist Sampling theorem.It can be combined data sampling and data compressing,which not only reduced the waste of sampling resources and easy to transport,but also reduces the pressure of data storage and the requirement to the hardware equipment.Therefore,this paper applies it to medical pathological image compression.On the basis of the study of block Compressed Sensing(BCS),this paper proposed a medical pathological image compression algorithm based on image blocks classification.First,edge image blocks are judged and classified using Canny edge detection operator and block image axial information,and non-edge image blocks are classified according to texture information.Then,to do a non-uniform sampling according to the results of the classification,which can reduce the waste of sampling resources to a certain extent.Finally,to training special sparse dictionary according to characteristics of various image blocks to improve the quality of the final reconstruction image.Experimental results show that the algorithm in this paper improved the image reconstruction,especially in the edge texture region.On this basis,the algorithm is further improved according to the characteristics of pathological medical images.To some block images,it is not adopting the processing mode of CS rather than random sampling.The experimental results show that,in the guarantee of the quality of reconstruction image,the image compression ratio is reduced by using the improved algorithm than before.This thesis also first explore the influence of color space for image reconstruction.By analyzing the correlation between the data of each channel and adding the color space transform to the proposed pathological medical image compression algorithm,this paper explored the medical pathological image compression sensing algorithm based on color space transformation.Through the color space conversion of the pathological medical color image,the correlation of each channel data is reduced,so that the reconstruction quality of the image is improved.The main research contents are as follows:In first chapter,mainly elaborated the research background of this paper,and then introduced the current research and development status of telemedicine system and Compressed Sensing theory.In second chapter,mainly elaborated the compressed sensing theory and its main components which are the selection design of the measurement matrix,the sparse representation of the signal and the reconstruction algorithm of the signal,and the each components are introduced in detail.In third chapter,the block Compressed Sensing is introduced,and on this basis it proposed a medical pathological image compression algorithm based on image blocks classification.It takes into account the structure of the image and its texture to improve the quality of the reconstruction image under a certain compression ratio.This algorithm use Matlab to simulate,and give analysis for the results of the experiment.And in the case of the quality of the reconstruction image,the algorithm is further improved to reduce the compression rate on the basis of the characteristics of medical images.In forth chapter,briefly introduce color space and its classification.Several common color space models and their conversion formulas with RGB color space are introduced in detail.Analysis the correlation of each channel data in each space model.Introduce that traditional processing method of color image compression perception.Explore the medical pathological image compression sensing algorithm based on color space transformation and give analysis for the results of the experiment.In fifth chapter,To summarize this article,and put forward the next step work.