An improved robust watershed for contour detection
Di Yang (RSISE)
COMPUTER VISION AND ROBOTICS SERIES PhD monitoringDATE: 2011-10-13
TIME: 14:15:00 - 14:45:00
LOCATION: RSISE Seminar Room, ground floor, building 115, cnr. North and Daley Roads, ANU
CONTACT: JavaScript must be enabled to display this email address.
ABSTRACT:
The watershed transformation was originally proposed by Digabel and Lantuejoul and later improved by Beucher and Lantujoulin in late of 70s as a tool for segmenting grayscale images. Nowadays it is used as an elemental step in many powerful segmentation procedures. The watershed constitutes one of the main concepts of Mathematical Morphology. The watershed transform can be classified as a region- based segmentation approach. Yet, original watershed is very sensitive to noise and unable to detect object with broken outline. These disadvantages severely influence on accuracy of image segmentation and contour detection, and limit its further implementation. The proposed method aims to make it immune to noise and expand its application domain to detect object with broken boundaries, with inappropriating surface tension coefficient into original algorithm.
BIO:
