Design And Implementation Of A Microfluidic Chip's Autofocus And Image Detection System

In this paper,starting from the micro drop type digital PCR equipment,according to the micro drop processing before PCR amplification,some sample droplet contains the nucleic acid target molecule to be detected,but some do not contain the nucleic acid target molecule to be detected.Through the laser-induced fluorescence detection method,the final concentration detection is realized.In this paper,the optical detection system of the micro drop digital PCR instrument is studied,in which the microfluidic chip is the target operation detection object in the experimental design.At present,some functions still need manual operation in micro drop type digital PCR equipment.Therefore,the automatic focusing of microfluidic chip and the realization of chip channel detection will be completed through the experimental design of this paper.The two problems solved can greatly improve the degree of automation and precision detection of the equipment.According to the theoretical knowledge learning and experimental design,the research content includes the following three aspects:1.Design and implementation of auto focus.In the research of Auto Focus module design,including image definition evaluation algorithm,focus search algorithm,focusing window design.The performance of image sharpness evaluation function is different in different actual environments.Therefore,according to the proposed several image sharpness evaluation functions,to give the real scene image data set and complete simulation through the MATLAB.According to the resent draw the sharpness evaluation curve.Finally get the most suitable evaluation functions in this system.In the study of focused search algorithm,three traditional search algorithms are introduced firstly,which the improved hill-climbing search algorithm is the best.In addition,this paper also proposes a search algorithm based on RBF neural network.Through the training and learning of the neural network,as long as input the fuzzy picture,the position information of the best focus will be output,so as to effectively reduce the frequent movement of the motor.Choosing a proper focusing window can effectively improve the efficiency of the algorithm.2.Design and implementation of image target detection.This paper mainly studies two detection methods,blob analysis and Hough transform line detection algorithm.After the realization of the actual video images collected by the two algorithms respectively,according to the comparison and analysis of the final detection results,the image target detection algorithm which is most suitable for the optical detection system in the experimental design of this paper is selected.3.Design of software control system of autofocus and image target detection technology.Based on the established experimental platform architecture,according to the requirements of the micro drop digital PCR system,select the appropriate hardware equipment to complete the design of the human-computer interface on PC.The functions that can be realized in the interface include auto focus,motor control and image target detection.Finally,the experimental results are analyzed and the improvement scheme is proposed.