Three-dimensional skull registration based on clifford algebra pupil distance invariability and visualization platform building

Aiming at the multi-modal three-dimensional skull registration, a methodology for constructing invariants of pupil distance relative features based on Clifford algebra is proposed in present work. Through hybrid programming of MATLAB and C#, a corresponding visualization registration platform is built for visualization operation. The algorithm proposed here is accomplished in the Clifford algebra space, in which binocular corneal highest feature points is utilized to calculate the pupil distance and construct geometric feature axis, and the registration is realized by rotation operator. The visualization registration platform integrates functions of three-dimensional reconstruction, translation, rotation, intercepting section, database access, etc. It allows researchers to analyze the registration fusion effect better. Additionally, MySQL database is embedded in the platform to store registration procedures, which provides intuitive and detailed registration information for medical workers and plays a great supporting role for the furthermore study of registration algorithm. Experimental results indicate that the registration algorithm is effective, and the visualization platform also performs well in medical image visualization and interactive operation.