Abstract:The market of handwriting recognition applications is increasing rapidly due to continuous advancement in OCR technology. This paper summarizes our recent efforts on offline handwritten Chinese script recognition using a segmentation-driven approach. We address two essential problems, namely isolated character recognition and establishment of the probabilistic segmentation model. To improve the isolated character recognition accuracy, we propose a heteroscedastic linear discriminant analysis algorithm to extract more discrimination information from original character features, and implement a minimum classification error learning scheme to optimize classifier parameters. In the segmentation stage, information from three different sources, namely geometric layout, character recognition confidence, and semantic model are integrated into a probabilistic framework to give the best script interpretation. Experimental results on postal address and bank check recognition have demonstrated the effectiveness of our proposed algorithms: A more than 80% correct recognition rate is achieved on 1,000 handwritten Chinese address items, and the recognition reliability of bank checks is largely improved after combining courtesy amount recognition result with legal amount recognition result. Some preliminary research work on Arabic script recognition is also shown.

A Novel Segmentation Method of Handwritten Chinese Amount Strings

Research on Segmentation of Unrestricted Handwritten Chinese Characters

An Effective Multi-stage Segmentation Method for Handwritten Chinese Characters

Handwritten Chinese Character Segmentation Using a Two-Stage Approach

A Multi-Stage Approach For Segmenting Handwritten Chinese Characters

A segmentation algorithm for handwritten Chinese character strings

Segmentation and Recognition System of Handwritten Chinese Bank Check Amounts

Segmentation And Recognition Of Continuous Handwriting Chinese Text

Multi-Step Segmentation Method Based on Adaptive Thresholds for Chinese Calligraphy Characters.

The Segmentation Algorithm for Handwritten Numeral Strings in Bank-Check Recognition

Study on Segmentation Algorithm for Unconstrained Handwritten Numeral Strings

Segmentation of On-Line Handwritten Chinese Characters Based on Stroke Merging and Dynamic Programming

Segmentation of Connected Handwritten Chinese Characters Based on Stroke Analysis and Background Thinning

Using MST in Handwritten Chinese Characters Segmentation

A Novel Segmentation and Recognition Algorithm for Chinese Handwritten Address Character Strings

A Two-Stage Online Handwritten Chinese Character Segmentation Algorithm Based on Dynamic Programming

Segmentation of Overlapped and Touching Handwritten Chinese Strings

A Distinguishing Method of Printed and Handwritten Legal Amount on Chinese Bank Check

Multi-step Segmentation Method Based on Minimum Weight Segmentation Path for Ancient Handwritten Chinese Character

Strategy for Segmentation of Legal Amount Images

Segmentation-Driven Offline Handwritten Chinese And Arabic Script Recognition