Position control using 2D-to-2D feature correspondences in vision guided cell micromanipulation

Area of Science:

• Microscopy and Micromanipulation
• Robotics and Automation
• Biotechnology

Background:

• Conventional camera calibration methods fail in micro-level cell operations due to hardware deviations and external disturbances.
• These deviations invalidate extrinsic camera parameters, impacting image processing and positional accuracy in biological micromanipulations.
• Existing macro-world visual servoing techniques often overlook these parameter invalidations, leading to errors in micro-scale tasks.

Purpose of the Study:

• To develop a novel algorithm for enhancing the success rate of vision-guided biological micromanipulations.
• To address the limitations of conventional camera calibration in micro-level cell operations.
• To enable simultaneous manipulator adjustment and position control during micromanipulation tasks.

Main Methods:

• A new algorithm was designed and implemented to monitor changing image patterns of micromanipulators (e.g., injection micropipette, cell holder).
• The algorithm utilizes 2-dimensional (2D)-to-2D feature correspondences for precise tracking.
• It incorporates real-time adjustment of the manipulator and simultaneous position control.

Main Results:

• The novel algorithm effectively monitors manipulator image patterns and detects deviations.
• It enables simultaneous adjustment of the manipulator and precise position control.
• When deviations occur, the system automatically retracts the manipulator to the initial focusing plane.

Conclusions:

• The developed algorithm significantly improves the accuracy and reliability of vision-guided biological micromanipulations.
• Real-time monitoring and correction of manipulator deviations are crucial for micro-level cell operations.
• This approach enhances the success rate of delicate biological tasks by maintaining positional integrity.