Character Recognition with Histogram Band Analysis of Encoded String and Neural Network
Parag S Deshpande, Latesh Malik, Sandhya Arora · 2005
This paper proposes a novel method of recognition of hand written character alphabets artificial neural networks together with conventional techniques. However in real situation it is not easy to build robust recognition because of vast variations in personal writing styles. There are also differences in one person’s writing style depending on the context, mood of writer and writing situation. The proposed method solves the problem by combining image encoding, histogram of encoded strings and training of neural network with the same information. This technique can be divided into five major steps: 1) Digitization of image 2) Thinning of binary image using a parallel thinning algorithm 3) Segmentation of the image 4) Extraction of features by encoding image in string form and find relationship between the structure of one segment to the structure of previous segment 5) Histogram band analysis of encoded string 6) Training of neural network with the information extracted from histogram band analysis and classify as a particular character using feed forward neural network trained by back propagation The authors have implemented a prototype system based upon the proposed method and conducted some experiments using the system. Experimental results support effectiveness of the proposed idea.