Research On Automatic Focusing Technology Of Micro-visual Measurement System

In the research of the autofocus based on the digital image processing,the image evaluation and search algorithm are the important part of this process.To achieve an accurate and fast automatic focusing of the system,selecting an evaluation algorithm with good evaluation performance and an efficient optimization search algorithm are the key issues that we need to solve urgently.This paper proposes an improved algorithm based on the "three-domain" image sharpness evaluation and an optimization algorithm for variable-step global search to solve the problems of low focusing success rate and slow focusing speed of the common algorithms.The automatic focusing technology of the micro-visual system is developed and implemented.A test platform to verify the effectiveness of autofocus algorithm is established.The main research contents of the paper are as follows:1.An autofocus scheme based on the active focusing depth method is proposed,and the hardware platform of the system is built based on the existing laboratory equipment.The contents include the selection of some hardware and the specific implementation of the entire platform system,which can lay a hardware foundation for the development of the entire system.2.The study of the image evaluation algorithms is carried out.By aiming at the problem of the misjudgment of gray-scale variation algorithms,a new algorithm of “three-neighborhood” image evaluation is proposed.The evaluation function curve figures of the proposed algorithm and the common gray-scale algorithms through experiments.The experimental results show that the proposed algorithm has the best evaluation performance and lays a theoretical foundation for the realization of the follow-up autofocus.3.For the problem of the slow focusing speed and the local extremum interference in the common autofocus search algorithms,an adaptive variablestep-length search algorithm and a four-step heuristics are proposed.Through the autofocus search process analysis,the autofocus experiment is performed on the platform by combing with the image evaluation algorithm.The experimental results show that the proposed algorithm can effectively reduce the number of the focusing search steps of the system and increase the focusing speed of the system.4.The autofocus experimental platform is set up to realize the development of the autofocus detection system.Through the focusing accuracy and measurement accuracy experiments,the automatic focusing algorithm is verified.The experimental results show that the proposed image evaluation algorithm and the search algorithm can achieve a rapid and accurate automatic focusing of the detection system,and the measurement accuracy can meet the system requirements.