A Cell Projection Algorithm Based on Feature Region Segmentation and Cells Sorting
Because of the complexity of the human structure, and the irregularity of the human tissues, and the discrepancy of human individuality, segmenting and rendering 3D medical data field are great challenge jobs. Considering the fuzzy characteristics of medical data field, the paper presents a novel sorted cell block projection algorithm based on feature region segmentation and cells sorting of 3D medical data field. Firstly the existence of the feature region ?edge transition region is demonstrated. Secondly, 3D data field is pretreated to a integral field, and the paper extends the method of segmenting 2D image and makes it can be used to segmenting 3D data field to transform traditional hard segmentation to edge transitional region segmentation which decreases segmentation difficulties, improves segmentation accuracy, and satisfies the requirements of the follow-up visualization process. Thirdly, According to the results of segmentation and the interesting grade of users, cells in the edge transition regions and the non_edge transition regions are combined with different coefficient to guarantee rendering image quality and decrease the number of the cells to be processed. Then, a new volumetric data structure and its storage structure are designed to sort cells according to their feature value, which makes the algorithm can select projected cells to decrease the number of the cells to be processed further. Finally the realization and experiment results show that the algorithm can improve image quality of edge transition region and improve rendering speed effectively.
Feature Region Segmentation Cells Sorting Cell Projection 3D Data Field
Xinwu Li
Electronic Business Department of Jiangxi University of Finance and Economics, Nanchang 330013, China
国际会议
上海
英文
1999-2003
2007-03-12（万方平台首次上网日期，不代表论文的发表时间）