A Threshold Selection Method for Caracter Images in a Document

Nakao Ichiro, Emura Satoshi, Takakura Saki, Takenouchi Mariko

Bibliographic Information

Other Title

文書中の文字画像に対する二値化閾値選択方式について

Description

Most of Character recognition methods for a document image is applied for binarized images.However,a more effective method is to input a document in gray scale and to select the optimal threshold for individual character areas in the document,thereby improving the precison of character recognition In this paper,we propose a method that selects the threshold for individual character areas by repeating binarization and extraction of character areas from a gray scale document image,and show that this method improves the quality of binarized character images and the precision of character recognition.Furthermore we propose the method to select the threshold for character images by estimating the black pixel ratio of original character fonts,and evaluate by character recognition ratio.

Journal

IEICE technical report. Artificial intelligence and knowledge-based processing

IEICE technical report. Artificial intelligence and knowledge-based processing 93 (317), 37-44, 1993

The Institute of Electronics, Information and Communication Engineers

Keywords

Details 詳細情報について

CRID: 1571698602336823808

NII Article ID: 110003186942

NII Book ID: AN10013061

Text Lang: ja

Data Source

CiNii Articles

Export

Report a problem