An improved matched filter for blood vessel detection of digital retinal images

  • Authors:

  • Mohammed Al-Rawi;Munib Qutaishat;Mohammed Arrar

  • Affiliations:

  • Computer Science Department, King Abdullah the Second School for Information Technology, University of Jordan, P.O. Box 13496, Amman 11942, Jordan;Computer Science Department, King Abdullah the Second School for Information Technology, University of Jordan, P.O. Box 13496, Amman 11942, Jordan;Computer Science Department, King Abdullah the Second School for Information Technology, University of Jordan, P.O. Box 13496, Amman 11942, Jordan

  • Venue:
  • Computers in Biology and Medicine
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

The matched filter has been widely used in the detection of blood vessels of the human retina digital image. In this paper, the matched filter response to the detection of blood vessels is increased by proposing better filter parameters. These filter parameters are found by using an optimization procedure on 20 retina images of the DRIVE database. Comparisons with other approaches show that the matched filter that uses the newly found parameters outperforms the matched filter that uses the classical filter parameters as well as some vessel detection techniques. A technique is also discussed to find the best threshold value for the continuous matched filter output image and hence the best segmented vessel image.