Multiple-cell recognition and path planning for robotic microinjection system

This paper presents a cell recognition method aiming at guiding the implementation more simply and effectively for injecting multiple cells using biological cell injection systems. The cells are randomly placed in the field of view of a microscope. In particular, the method is proposed to guide the cell grasping device through computer vision to grasp multiple cells in a field of image region without the need for any preparatory work. A coarse-to-fine pattern matching method is used to identify multiple cells. The path planning is realized by two optimization approaches in terms of genetic algorithm and closest segment insertion method. The Vision Development Module of LabVIEW is used with the Integrated Development Environment for implementing this method. Experimental studies are conducted to verify the effectiveness of the proposed approach. The reported method will automate the cell injection process to a greater extent.