A LEXICON DIRECTED ALGORITHM FOR RECOGNITION OF UNCONSTRAINED HANDWRITTEN WORDS

In this paper, authors discuss on a lexicon directed algorithm for rec ognition of unconstrained handwritten words (cursive, discrete, or mix ed) such as those encountered in mail pieces. The procedure consists o f binarization, presegmentation, intermediate feature extraction, segm entation recognition, and post-processing. The segmentation recognitio n and the post-processing are repeated for all lexicon words while the binarization to the intermediate feature extraction are applied once for an input word. This algorithm is essentially non hierarchical in c haracter segmentation and recognition which are performed in a single segmentation recognition process. The result of performance evaluation using large handwritten address block database, and algorithm improve ments are described and discussed to achieve higher recognition accura cy and speed. Experimental studies with about 3000 word images indicat e that overall accuracy in the range of 91% to 98% depending on the si ze of the lexicon (assumed to contain correct word) are achievable wit h the processing speed of 20 to 30 word per minute on typical work sta tion.