Research And Design Of Wireless Optogenetics Experiment System Based On Particle Filter Tracking

Optogenetics technique is an emerging neuromodulation method with the advantages of precise targeting,rapid response,and so on.The implanted fiber optic method used in the traditional optogenetic technique method limits the activities of the objects and has a certain impact on animal behavior analysis.Therefore,the wireless research of optogenetic experiment system based on visual technology for position and tracking is of great significance to the practical value.In view of the shortcomings of most of the mainstream optogenetic experiment system,this thesis proposes and designs a wireless optogenetic experiment system based on visual tracking without load-bearing,non-implantation and satisfying large-scale irradiation.According to the analysis of the mainstream visual tracking algorithm,apply the particle filter algorithm with built-in state transition model to experiment system.Particle filter firstly samples particles in prediction region,then obtains the weight of each particle combining observation model.After iterative calculation,the posterior probability density of the state vector is obtained to finish tracking.The traditional particle filter optimization strategy mostly adopts the feature fusion method,which leads to the increase of the dimension of observation model and computational complexity and the lower sampling efficiency.Due to the problems in the observation model and the proposal sampling,an improved particle filter tracking algorithm was proposed,which includes background discriminative feature selection and the batch proposal sampling method.Among them,the selection of the representative observation features based on comprehensive information of target and background improves the adaptive ability of the particle filter observation model;and batch proposal sampling based on state transition equation and weight feedback theory improves the diversity of sampling particles while reducing the variance of the sampled particle distribution,and suppresses particle degeneracy and particle starvation.According to the specific requirements of optogenetic experiment,the thesis elaborates the architecture and design scheme of experiment system including video tracking,automatic control modules and so on,and completes the design and verification of each module.In the video tracking module,combined with the state transition model of optogenetic mice,instantiate application of the improved algorithm and analyze the position difference between the mouse and the irradiation point based on the target imaging model;in the automatic control module,segmented PD control is adopted to drive the stepping motor to avoid lost-step and low frequency oscillation problem of stepping motor,which improves the stability of the system and realizes the closed-loop control of the system.Finally,the thesis tests the improved particle filter algorithm,module elements of system and system physical platform.The experiment results show that the improved particle filter algorithm is superior to the traditional feature fusion method and the optogenetic experiment system based on particle filter tracking has a high application value,which provides a novel solution for optogenetic experiment and conduces to the development and progress of optogenetic technology.