Recognition Motivation • Today device that interact with our computer are keyboard, mouse, etc. • They provide us with different inputs which are then utilize by various application for operational use • Basic GSM phones -> J2me -> Symbian -> android -> ios • Traditional style navigational keyboard -> qwerty -> touch Need for project • This project stresses over development of a new way of interaction between users and computers. • It is a modular system which will utilize any bright source as the input which is detected by the camera mounted on the computer screen. Intro • OCR is the recognition of the optical characters on the basis of image processing and correlation of the processed image with the database. • We used alphabets as the input methods so that a large number of inputs can be provided. • Low light room. • Bright light room. • Room with white objects. • Room having many windows or having many source of natural light Principal of project • There is a feature that any specific function can be defined for a specific alphabet in the algorithm such that whenever that alphabet will be detected the function assigned to it will be executed. • The alphabet detected is then sent to a python script which can perform many applications like: • will search all the files and folders starting with that alphabet in the current location. • will print the current opened document • will start the slide show for current powerpoint presentation Work Flow Working • Why only White Light? • Y=235, Cb=128, Cr = 128 • Easy to recognize. Minimal interference with other colours. • Tried with Red light (projector pointer) : • Y ~ 65 , Cb ~ 100 , Cr ~ 212 • High intensity source, but less surface area • Interference with surroundings Simulation and testing • The project was run under different conditions. The basic purpose of it was to check the results in different conditions and to see the probability of error. The different conditions used for the experiment are as follows: • Low light room. • Bright light room. • Room with white objects. • Room having many windows or having many source of natural light Result and constraints • Low light room: the project performed best without any requirement of any additional camera setting • Bright light room: the project faced many difficulties and worked with error • Room with white object: a little noise reduction technique eliminated all the error Practical use of project / Application area • The user can can open the application in various scenario to make its document look more sophisticated with this ocr detection on text Eg. • Presentation • Code editor • File manager • desktop Future Plan • This project stresses over development of a new way of interaction between users and computers. • It is a modular system which will utilize any bright source as the input which is detected by the camera mounted on the computer screen. • The OCR algorithm to our project will detect the alphabet written by the user on-air. • It will also perform a predefined function inside computer accordingly like opening browser (for character ‘E’) or any specific application for a specific alphabet. biblography • https://web.stanford.edu/~boyd/cvxbook/ • Bishop, Pattern Recognition and Machine Learning, 2006 • Ripley, Pattern Recognition and Neural Networks, 1996 •