Research Of Volume Rendering And GPU Acceleration Based On Medical Image Sequence

The main content of this paper study is the volume rendering algorithm in the3D medical image reconstruction, and specific research of the paper is the ray casting method and its principle in accelerating technology, on the basis of the research the volume rendering algorithm’s GPU acceleration method.3D medical image reconstruction technology is one of the popular research direction in the medical application research, it can make image and graphic processing technology apply to medical diagnosis and treatment, but also promote to the remote medical treatment, if3D reconstruction technology is mature, it will be widely used in business, and will promote the global sharing of medical technology and the widespread use of remote medical treatment, this will greatly improve people’s health level, so the3D reconstruction technology of the medical image has important practical significance.Now3D medical image reconstruction technology research is still in the theory and clinical application research. To realize real a wide range of commercial use, we need constantly strive to study.Before the3D medical image reconstruction technology, firstly we should get medical data and make these data some pretreatment, this paper mainly introduces two kinds of medical image such as the CT images and MRI image, and their characteristics and principle of produce to carry on the analysis, but also make the two images compared. Then introduces the DICOM communication standards and its part. Before the reconstruction image, to get medical data image filtering and the place such as image enhancement Richard. Image filtering is to eliminate the noise in the data, and image enhancement is in order to eliminate the other information in the process of strengthening the region of interest, make these useful information strengthened, so researchers will get higher image quality. In a word, the pretreatment technology is for the following3D reconstruction process to do a good job in basic.In3D reconstruction algorithms are mainly two kinds of algorithm of surface rendering and volume rendering, the use of surface rendering reconstruction algorithm can only draw the surface of the image, and can’t show the image of the internal information, but it has the advantage of fast calculation speed; And the use of volume rendering reconstruction of advantage of the algorithm is able to display3D images of internal information, and it have a good practical value. Because of drawing a lot of data, its disadvantages is drawing speed slow, cannot satisfy the requirement of real-time rendering medicine, so we can through some acceleration technology to improve volume rendering algorithm. This paper mainly studies the volume rendering3D reconstruction algorithm, in the volume rendering reconstruction algorithm, mainly includes ray casting method, shear-deformation method, sputtering method, transform domain method and3D texture mapping based on hardware. All the volume rendering method has its own advantages. The ray casting method in this paper is the main algorithm, the ray casting algorithm get quality the most high in this methods, of course, are slower, this paper use VTK software tools SDK to realize the ray casting method. In the traditional ray casting method, on the basis of this paper puts forward a new improved algorithm to accelerate ray casting method, the new improved algorithm is combining box accelerating technology and close cloud algorithm based on space jump together new acceleration algorithm, the experimental results show that the new algorithm can quickly improve the traditional ray casting algorithm, but also on the quality is basically remain the same.Finally, the ray casting method of GPU acceleration technology is studied in this paper, using CUDA technology to the traditional ray casting method and the new accelerated algorithm, and CUDA is refers to the unified computing equipment structure, can use GPU to hardware program for parallel processing, and ray casting algorithm in the calculation of each ray is separate, so ray casting method is very suitable for parallel processing. The experimental results show that the new accelerated algorithm using CUDA technology can greatly improve the speed of the algorithm, basically has reached the requirement of real-time rendering.