A Handwritten Numerals Recognition Method Based On Skeleton Structural Features
Luo Zhen-x · Microcomputer Information · 2010
In this paper, a new method based on skeleton structure features for recognition of handwritten numerals is proposed. Use the numerals skeleton after thinning and other preprocessing, and according search matching, extracted digital skeleton endpoint and cross-point structure features and then basis the location characteristics of the endpoint for the digital identification. The algorithm is easy to implement, has a high reliability and recognition rate, can meet the handwriting recognition applications. This method has been successfully applied in forms automatic entry and identification system.