Contrast enhancement and clustering segmentation of gray level images with quantitative information evaluation

  • Authors:

  • Zhengmao Ye;Habib Mohamadian;Su-Seng Pang;Sitharama Iyengar

  • Affiliations:

  • Southern University, Baton Rouge, LA;Southern University, Baton Rouge, LA;Louisiana State University, Baton Rouge, Louisiana;Louisiana State University, Baton Rouge, Louisiana

  • Venue:
  • WSEAS Transactions on Information Science and Applications
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Improper illumination and medium dispersing could occur in quite some gray level image collecting processes. Contrast enhancement and clustering segmentation are two effective approaches for the related pattern recognition problems. Image enhancement and image segmentation can be applied to different areas of science and engineering, such as biometric identification, national defense and resource exploration. Thus, adaptive image enhancement can be implemented to improve the image quality and to reduce random noise simultaneously, which is used to adapt to the intensity distribution within an image. Nonlinear K-means clustering can be applied for image segmentation, which is to classify an image into parts which have strong correlations with objects in order to reflect the actual information being collected. For example, it can be used against the effects from unevenly distributed pressure or temperature condition under atmosphere medium and water medium. To evaluate the actual roles of image enhancement and image segmentation, some quantity measures should be taken into account. In this study, a set of quantitative measures is proposed to evaluate the information flow of gray level image processing. Concepts of the gray level energy, discrete entropy, relative entropy and mutual information are proposed to measure outcomes of the adaptive image enhancement and K-means image clustering.