Study on the automatic multi-cells tracing algorithm in a sequence of micro images

The varying density of the cell culture and the complexity of the cell behavior, shape deformation, mitosis, close contact and partial occlusion pose many challenges to existing tracking techniques. In this paper, a novel automatic multi-cells tracing algorithm is put forward. We implemented the automatic tracking method that proceeds in three steps: Firstly, we detect the individual cells in each frame using feature point detection techniques. Secondly, we establish the correspondences between cells over neighboring frames with a notion of decision risk derived from cell motion priors and record them in the track linked list data model. Finally, the multi-cells' trajectories are formed from the track linked list data model. The experimental result shows that the multi-cell tracking algorithm simultaneously tracks a very large number of cells and can successfully cope with the aforementioned challenges.