Character recognition from image using radial basis function neural
Images contain different sorts of helpful data that ought to be removed at whatever point required and this data might be as content present in image. Character Recognition in Images has turned into a potential application in many fields like Image ordering, Robotics, Intelligent transport frameworks, etc. The main aim of this work is to create an optical character recognition software that extracts characters from image using Radial Basis Function Neural Networks (RBFNN) and develop an intelligent classifier system that recognizes and classifies text characters from images. The application is able to recognize characters online and offline. A test pad was created in the form of a drawing board, it converts the drawn character into an image to check if it can be recognized and the predicted character is displayed (online recognition). PredictCharacter.m is used for offline recognition; this function gives a list of possible characters with percentages and the one with the highest percentage is the answer. The acknowledgment rate of Radial Basis Function Neural Network (RBFNN) is observed to be best as the recognition rate in the proposed framework lies in the range of 95.5 and 97%.
Keywords: Images, Character Recognition, Pattern Recognition, RBFNN