Another Method of Building 2D Entropy to Realize Automatic Segmentation

摘要：

2D entropy formed during the process of building 2D histogram can realize automatic segmentation. Traditional method utilizes central pixel grey value and the others or all of pixels grey mean value in 4-neighbor to build 2D histogram. In fact, the change of the greyscale value between two invariable position vectors cannot represent the total characteristics among neighbour pixels very well. A new method is proposed which makes use of minimum grey value in the 4-neighbor and of maximum grey value in the 3×3 neighbour except pixels of the 4-neighbor. New method and traditional one are used in contrast to realize image automatic segmentation. The experimental results of the classical image prove the new method is effective.

作者: Y F Zhang Y Zhang

作者单位: Department of Mathematics, Harbin Institute of Technology, Harbin 150001, China;Department of inform Department of information Engineering, Harbin Institute of Technology, Harbin 150001, China

会议类型: 国际会议

会议名称: 第四届仪器科学与技术国际会议( 4th International Symposium on Instrumentation and Science and Tcchnology)

会议地点: 哈尔滨

会议语种:英文

页码: 303-307

在线出版日期: 2006-08-08（万方平台首次上网日期，不代表论文的发表时间）