Abstract
Contour representations are well suited for representing handwritten word images. The storage is lossless and the processing time required is a relatively small. While chain code encoding of contours has been favored by researchers in the area of handwriting recognition the representation does have some drawbacks. Firstly, a simple shift of a single pixel can cause a break in the contour, or introduce a point that does not adhere to the property that the contour must be a single pixel wide at all points of the contour. Such an event can occur during slant correction or noise removal. the speed efficiency of a character recognizer that uses chain code representation, For lexicons of size 1000, the time consumed by the chain code based preprocessing and feature extraction routines is less than 10% of the total time.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
