Y.K. Chung

Electronics and Telecommunications Research Institute, Sŏul, Seoul, South Korea

Are you Y.K. Chung?

Claim your profile

Publications (3)0 Total impact

  • Y.J. Song · K.C. Kim · Y.W. Choi · H.R. Byun · S.H. Kim · S.Y. Chi · D.K. Jang · Y.K. Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we propose a text extraction method from camera-captured document style images and propose a text segmentation method based on a color clustering method. The proposed extraction method detects text regions from the images using two low-level image features and verifies the regions through a high-level text stroke feature. The two level features are combined hierarchically. The low-level features are intensity variation and color variance. And, we use text strokes as a high-level feature using multi-resolution wavelet transforms on local image areas. The stroke feature vector is an input to a SVM (support vector machine) for verification, when needed. The proposed text segmentation method uses color clustering to the extracted text regions. We improved K-means clustering method and it selects K and initial seed values automatically. We tested the proposed methods with various document style images captured by three different cameras. We confirmed that the extraction rates are good enough to be used in real-life applications.
    No preview · Conference Paper · Jan 2005
  • Y. K. Chung · S. Y. Chi · K. S. Bae · K. K. Kim · D. Jang · K. C. Kim · Y. W. Choi
    [Show abstract] [Hide abstract]
    ABSTRACT: When document images are obtained from digital cameras, many imaging problems have to be solved for better extraction of characters from the images. Variation of illumination intensity sensitively affects to color values. A simple colored document image could be converted to a monochrome image by a traditional method and then a binarization algorithm is used. But this method is not stably working to the variation of illumination because sensitivity of colors to variation of illumination. For narrowly distributed colors, the conversion is not working well. Secondly, in case that the number of colors is more than two, it is not easy to figure out which color is for character and which others are for background. This paper discusses about an extraction method from a colored document image using a color process algorithm based on characteristics of color features. Variation of intensities and color distribution are used to classify character areas and background areas. A document image is segmented into several color groups and similar color groups are merged. In final step, only two colored groups are left for the character and background. The extracted character areas from the document images are entered into optical character recognition system. This method solves a color problem, which comes from traditional scanner based OCR systems. This paper also describes the OCR system for character conversion of a colored document image. Our method is working for the colored document images of cellular phones and digital cameras in real world.
    No preview · Article · Jan 2005
  • Source
    K.C. Kim · H.R. Byun · Y.J. Song · Y.W. Choi · S.Y. Chi · K.K. Kim · Y.K. Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: We propose a method that extracts text regions in natural scene images using low-level image features and that verifies the extracted regions through a high-level text stroke feature. Then the two level features are combined hierarchically. The low-level features are color continuity, gray-level variation and color variance. The color continuity is used since most of the characters in a text region have the same color, and the gray-level variation is used since the text strokes are distinctive to the background in their gray-level values. Also, the color variance is used since the text strokes are distinctive in their colors to the background, and this value is more sensitive than the gray-level variations. As a high level feature, text stroke is examined using multi-resolution wavelet transforms on local image areas and the feature vector is input to a SVM (support vector machine) for verification. We tested the proposed method with various kinds of the natural scene images and confirmed that extraction rates are high even in complex images.
    Preview · Conference Paper · Sep 2004