Title :
Cells tracking based on camshift and active contour model
Author :
Liu Ningxia ; Ding Xianting ; Sun Mingzhu ; Zhao Xin ; Lu Guizhang
Author_Institution :
Inst. of Robot. & Autom. Inf. Syst., Nankai Univ., Tianjin, China
Abstract :
A WPR-Camshift automated cells tracking algorithm is proposed here to solve shortcomings in the existing algorithms such as easy losses of tracking fast-moving cells, inadequate information of cell shape. By means of weighted probability distribution and prediction of moving cells´ locations, the algorithm can manage to track fast-moving cells, meanwhile, precise locations and shapes of cells can be achieved through re-locating cells and extracting the cell boundaries with active contour model. Besides, estimation of the critical state of cell division, segmentation and tracking of daughter cells just after division are included in the algorithm. The effectiveness of the method has been verified by testing in 3 groups of cell pictures which conclude 135 cells. Experimental results show that the algorithm is able to track deformable, migrating cells and handle cell division, which is of comparatively high stability and accuracy.
Keywords :
biology; cellular biophysics; state estimation; statistical distributions; WPR-camshift automated cell tracking algorithm; active contour model; cell boundary extraction; cell division; cell pictures; cell relocation; cell shape; continuously adaptive mean-shift; critical state estimation; daughter cell segmentation; daughter cell tracking; fast-moving cells; moving cell location prediction; weighted probability distribution; weighted-predicted-relocated camshift; Active contours; Electronic mail; Prediction algorithms; Shape; Sun; Tracking; Camshift; cells tracking; motion prediction; weighted probability;
Conference_Titel :
Control Conference (CCC), 2012 31st Chinese
Conference_Location :
Hefei
Print_ISBN :
978-1-4673-2581-3