Cell Rotational Motion Intelligent Perception under Micro-Nano Scale Vision

摘要

Cell manipulation in drug research and development, physiology and other biomedical research plays an important role. In high-precision micro environment, especially when the measurement accuracy reaches the nanometer level, it will be subject to light diffraction, scattering and other issues of interference. Cell images in different target is characterized by high similarity, and the movement of cells in some specific conditions is extremely fast, especially in the rotation of the movement is difficult to use the traditional macro-visual algorithm to identify the measurement. In this paper, the SURF feature of the extracted image is selected and its feature registration algorithm is improved for the high-speed motion characteristics of the nanometer visual image are further marked. It mainly includes three parts: feature point detection, feature point description, feature registration and labeling. Finally, the validity of the algorithm is verified by experiments in high-speed rotational motion video.