Design And Development Of Intelligent Video Magnifier Software Based On Linux

Recently,low vision,which harms patients' physiological and psychological health,is becoming a focus of the public,and also,it always harms the patient's family.How to utilize the residual vision of the low vision patient effectively and to achieve visual improvement with medical or other methods has been an urgent problem that needs researchers to solve.Currently,kinds of visual assistant devices can be found in the market,and there are some video magnifiers for reading texts.But these devices are lack of operational convenience and have too few useful functions.Specially,these devices cannot meet the new requirements of users,such as enlarging the words without distortion,words scrolling display,etc.For tackling above problems,this thesis designs and develops a new intelligent video magnifier software which can help low vision patient to get better reading experience.The main work and achievements are presented as follows:(1)This thesis illustrates the background and the significance of this subject,analyzes the domestic and foreign research condition of electronic visual aids and the two used techniques(the optical character recognition as well as the vector font and engine),summarizes the main research content and organization structure of the thesis.(2)In order to increase the recognition ability,a processing method to text image is presented,including choice of optimal image and graying,tilt correcting,binaryzation,noise elimination of target contents.In order to extract the characters with format information in the image,this thesis studies and realizes the Tesseract OCR-based character recognition method.The experimental result shows that the effect of character recognition in this module can basically meet the requirement.(3)This thesis designs an image specification parameter management module which is responsible for managing the needed parameter in image reconstruction and display.This thesis designs and realizes an image reconstruction method based on FreeType font engine to change the character string into an image,and an effective image display method which makes color change,scrolling display and row spacing adjustment possible.(4)This thesis designs a GUI software which assembles the modules mentioned above based on Qt development framework,with these functions of importing the collected image,scaling image,controlling words reading,changing the color mode,adjusting row space,etc.(5)A summary of this thesis is made at last,and the prospect of the future development is given.