Research On Automatic Focusing Algorithm Of Lens Module Based On Image Processing

In the production process of prime camera,the lens module needs to be focused at the optical focus to flow to the next production process.However,the production mode of lens module prime is usually manual focus,which requires human eyes to observe whether the image is clear,adjust the lens to be clear,and then fix it with glue.In order to improve the production efficiency,the auto focus technology is emerging,but the existing definition evaluation method has better performance,poor applicability and stability for specific images.Therefore,this paper studies the methods and classification of the commonly used auto-focus technology,analyzes the contrast image sharpness evaluation algorithm,selects the algorithm suitable for the lens module and improves it,and discusses the optimization and improvement of the search algorithm.The main research work and methods are summarized as follows:(1)Automatic focus lens module theory,through the introduction of the composition of the optical lens module and cooperate with the imaging theory of geometrical optics model,this paper expounds the focus,focus,depth of focus,dispersion,depth of field,aperture,and the view Angle of the relationship between each other,and briefly describes the commonly used classification and method of automatic focus is based on image processing of automatic focusing system key points for the image clarity evaluation algorithm and search algorithm;(2)Image clarity evaluation algorithm research: this chapter first analyzes the algorithm of clarity evaluation function index and quantitative standard,and then choose a lens module collected image sequences on two targets,with the aid of image definition evaluation function is commonly used matlab simulation evaluation curve of target image and quantitative analysis,the results show that in this group of lens module,algorithm based on space domain was superior to other algorithms,the algorithm based on frequency domain and wavelet transform poor real-time performance,based on the statistics and information entropy algorithm does not apply;(3)Image sharpness evaluation algorithm improvement: this chapter of image gray level transformation and gaussian filter in the first place,the window of the multi regional method together with the weight selection method,designed the template definition evaluation algorithm based on improved sobel gradient,through theacquisition sequence images of different lens module,according to the definition evaluation index quantitative analysis contrast improvement algorithm and traditional algorithm,the results show that the improved algorithm is superior to the traditional algorithm,and has better timeliness;(4)Focusing algorithm analysis and optimization: this chapter summarizes the analysis of global traversal search algorithm,Fibonacci pole searching algorithm,curve fitting and ordinary mountain climbing algorithm principle and the advantages and disadvantages,on the basis of common climbing algorithm,combined with focus on stability,focusing efficiency and accuracy are the focal and versatility this algorithm evaluation factors,design a kind of based on image contrast method for optical lens module focus search algorithm,this algorithm has more phase step length automatic adjustment,direction judgment,is focused on the environment to determine the characteristics of stability,improve the stability of the focused search and quickness.