Automatic vision-guided robotic cell micromanipulation system

Abstract

In life science research and applications such as cell mechanobiology, cell micromanipulation is skill-intensive, involving cell immobilization, micropipette positioning, cell biomembrane deformation and penetration, etc. The imprecise and inconsistent positioning in manual cell micromanipulation and biomembrane stress control demands a vision-guided robotic system to achieve high-accuracy static position control and dynamic motion tracking. This dissertation emphasizes the various techniques in realizing the aforementioned system.

Calibration of a vision system is necessary to ensure accurate coordinate systems transformation. Automatic self-calibration without the aid of calibration templates is proposed to achieve on-line calibration for real-time applications. The difference between the measured output and computed output after calibration is less than the physical resolution of the vision system.