Automated tracking of single nano-particle in live cell

Automated tracking of moving particles in sequential frames can provide important information in live cell, such as subcellular structures and dynamical working mechanisms. A real-time scheme for tracking single nano-particle in live-cell frames acquired from microscope is proposed in this article. The scheme consists of a detection stage based on local searching and a global optimized matching stage. In the first stage nano-particles are detected and segmented by gray value morphology, and the searching is restricted in a limited region so as to promote efficiency of the execution. And then the global optimal matching is exploited to recognize the target nano-particle based on gray and spatial features in the second stage. Additionally, because the spurious disturbance from noise and other non-target nano-particles are eliminated, the scheme is adaptable to overlapping nano-particles, and also does well in recognizing the target in the situation of high nano-particle density.