Abstract:Offline Chinese handwritten character string recognition is one of the most important research fields in pattern recognition. Due to the free writing style, large variability in character shapes and different geometric characteristics, Chinese handwritten character string recognition is a challenging problem to deal with. However, among the current methods over-segmentation and merging method which integrates geometric information, character recognition information and contextual information, shows a promising result. It is found experimentally that a large part of errors are segmentation error and mainly occur around non-Chinese characters. In a Chinese character string, there are not only wide characters namely Chinese characters, but also narrow characters like digits and letters of the alphabet. The segmentation error is mainly caused by uniform geometric model imposed on all segmented candidate characters. To solve this problem, post processing is employed to improve recognition accuracy of narrow characters. On one hand, multi-geometric models are established for wide characters and narrow characters respectively. Under multi-geometric models narrow characters are not prone to be merged. On the other hand, top rank recognition results of candidate paths are integrated to boost final recognition of narrow characters. The post processing method is investigated on two datasets, in total 1405 handwritten address strings. The wide character recognition accuracy has been improved lightly and narrow character recognition accuracy has been increased up by 10.41% and 10.03% respectively. It indicates that the post processing method is effective to improve recognition accuracy of narrow characters.

Advanced Topics in Character Recongition and Document Analysis: Research Works in Intelligent Image and Document Research Lab, Tsinghua University

Chinese Calligraphic Style Representation for Recognition

Multilingual document recognition research and its application in China

Design And Development Of An Ancient Chinese Document Recognition System

SH-MQDF based calligraphic character recognition.

The Advances of Handwritten Chinese Character Recognition and Its Application in Multimodal Interface

Th-Ocr: System For Multilingual Document Analysis, Recognition And Reconstruction

Segmentation-Driven Offline Handwritten Chinese And Arabic Script Recognition

Advances in Character Recognition

A Deep Learning Technology based OCR Framework for Recognition Handwritten Expression and Text

A Novel Algorithm For Handwritten Chinese Character Recognition

From Textline to Paragraph: A Promising Practice for Chinese Text Recognition

An approach for handwritten Chinese text recognition unifying character segmentation and recognition

Dual-Branch Multitask Fusion Network for Offline Chinese Writer Identification.

Evaluation and application of recognition confidence in OCR

Post Processing for Offline Chinese Handwritten Character String Recognition

Hierarchical Approximate Matching for Retrieval of Chinese Historical Calligraphy Character

Handwritten Recognition Techniques: A Comprehensive Review

Lsh-Based Large Scale Chinese Calligraphic Character Recognition

Uyghur, Chinese and English Multilingual Document Recognition

Chinese Character Recognition with Augmented Character Profile Matching