Blackboard OCR

Abstract

BlackboardOCR is the automated recognition program for hand drawn inputs using chalks on Blackboard or whiteboard marker on Whiteboard. It can detect and recognise general geometric primitives and other forms of entity which is defined by the developers. This project aims to examine and cultivate the techniques of image processing and implement a program that could process an image of a diagram through Python OpenCV programming. The program can detect and recognise shapes and words. For the word detection, there are two ways of approaches, and that is for the program to detect on its own and for the user to specify where the words are. This report will include the coding scheme to accord with object detection, the different stages of implementation and present experimental results based on the accuracy of the processing of the image.