会议专题

Neuron stem cells tracking in image sequence based on Hungarian algorithm

Study on cells tracking is an essential part of cytology and biology and is of crucial importance in the study of drug and disease research.And the multitarget tracking in cell image sequences is a main difficulty in cells locomotion study.In this paper,tracking by detection method has been adopted.After segmentation of level set combined with average gray threshold in the whole image sequence has been accomplished,Hungarian algorithm is then applied in inter-frame stem cells matching.To the unmatched cells,which is classified as lost cells and emergence cells,we fulfill matching according to cells some biological characteristics.The improved algorithm has been tested in a rather low contrast ratio sequence.Experimental results show that our approach has improved tracking accuracy up to 12.5% compared with Zhangs algorithm based on topological constraints.And the final tracking accuracy can reach 81.25%.

image sequences neuron stem cells tracking segmentation Hungarian algorithm matching

Shasha Dong Chunming Tang Yucui Liu Ying Cui

Information and Communication Engineering college Harbin Engineering University Harbin,China

国际会议

2010 4th International Conference on Intelligent Information Techonlogy Application(第四届智能信息技术应用国际学术研讨会 IITA 2010)

秦皇岛

英文

252-255

2010-11-05(万方平台首次上网日期,不代表论文的发表时间)