Recognizing Overlapping Hand-Printed Characters by Centered-Object Integrated Segmentation and Recognition

This paper describes an approach, called centered object integrated seg(cid:173) mentation and recognition (COISR). The application is hand-printed character recognition. One uses a backpropagation network that scans exhaus(cid:173) tively over a field of characters and is trained to recognize whether it is centered over a single character or between characters. When it is centered over a character, the net classifies the cnaracter. The approach is tested on a dataset of hand-printed digits.