[Show abstract][Hide abstract] ABSTRACT: A novel segmentation algorithm for color video sequences using the
level set technique is proposed. This algorithm is applied to the
problem of automatic face region segmentation in video sequences. Given
a target color range training set, a model of that color distribution is
formed. An objective-function is then defined that seeks the boundaries
of any regions in an image that match the target colors. This function
is minimized using an implementation of the level set algorithm. Motion
estimation greatly reduces the computation required for segmenting video
sequences. Results are presented demonstrating the high level of
accuracy and flexibility of the proposed segmentation algorithm
Image Processing, 2000. Proceedings. 2000 International Conference on; 02/2000
[Show abstract][Hide abstract] ABSTRACT: The methods of active contours (snakes) and level sets were applied to images of the retina in order to locate the outer boundary of the optic disk. A gradient-vector-flow based active contour was used as it performed well over a large range of initial conditions. Images were pre-processed to lessen the influence of blood vessels on boundary detection. Both active contours and level set methods accurately located the correct boundary; level set methods were computationally more intensive.
[Show abstract][Hide abstract] ABSTRACT: In this paper, the level set algorithm is examined with a view to applications in image and video processing, and machine vision. Conventional parametric deformable models, which can be viewed as Lagrangian geometric models, where the boundary of the model has the ability to align itself with certain image features, are initially pre-sented. This paper describes some of the problems associ-ated with this approach and introduces the level set method as a step towards a solution. The main idea is to take the Lagrangian form of the parametric deformable model, and replace it with an Eulerian, initial value partial differential equation. This involves recasting the problem in a higher dimension, but it can be shown that this actually leads to a reduction in complexity. Various improvements to the basic form are presented, as well as several different formulations of the guiding speed function. For this paper, the level set method was applied to two areas; finding and tracking a human face in video-conferencing sequences, and extract-ing subtle contours from medical images. In applications where off-line computation is acceptable, the level set algo-rithm competes well with established techniques, and can be demonstrated to have several advantages over them. A disadvantage of the level set method over active contours has been found to be the high computational cost.