Visual servoing methods in robot-assist cell microinjection system

Robot-assist microinjection is attracting a lot attention for its potential in providing high throughput and high precision in biological industry. This paper addresses two visual servoing issues in a robot-assist zebrafish embryo microinjection system. The first one is the microscope autofocusing method with the aim of reducing the sophistication of the whole manipulation system. The suitability of different focus criteria are evaluated in the scenario of robotic embryo microinjection system and an optimization technique is proposed to speed up the autofocusing process. The other issue is the implementation of the visual processing algorithm for cell detection and location. Three different segmentation methods for cell detection and location are evaluated in the processing speed and precision.