An Approach to GUI Identification for Printed Gurumukhi and
Inderpreet Kaur, Saurabh Mahajan · 2015
Optical Character Recognition system is used to recognize printed and handwritten alphanumeric text from input image. A numerous of methods have been published based on optical character recognition. In proposed work expansion of optical character recognition to recognize multi-scripts is done which in infancy. Such type of expansion is crucial in India where each state has diverse language. The planned work is done to recognize Gurumukhi and English text only. Arial font is consider for English language and Gurbanikalmi font is consider for Gurumukhi language for the training of the proposed system and the recognition of image text evaluates with respect to following various stages; pre-processing, segmentation, feature extraction and classification are used to process the assumed image. The text of assumed image is segmented in lines, words and characters using proposed segmentation scheme after its processing through preprocessing stage. Then histogram projection profiles and number of holes features are calculated to recognize text. A GUI has been proposed to process the assume image and to calculate the recognition accuracy of system for different set of test samples. The end result of proposed system offers high accuracy and also provided information about correct and wrong recognized number of samples.