Samuel G. Burgiss
Electrical Engineering
M.S. 1998

 

 Range Image Segmentation through Pattern Analysis of Multi-Scale Wavelet Transform

Research Objectives:
This research focuses on segmenting range images using multi-scale edge information. Image segmentation is a difficult problem, but it is an essential pre-processing step in many computer vision systems. In the case of range images, segmentation is performed in order to locate objects, fit surfaces, and render volumes. Segmentation of a range image is the process of dividing the image into areas that are associated with specific objects or object components. Often, these objects or object components can be identified in a range image as patches that are relatively uniform in range value or surface shape.

Methodology and Results:
This work presents an image segmentation method for range data that uses multi-scale wavelet analysis in combination with pattern recognition. To segment range images, a technique referred to as pattern analysis of scale space for the detection of features (PASSEF) has been developed. Then, a morphological watershed algorithm is applied to a fuzzy edge map to implement segmentation. The system uses pattern recognition to classify points in an image based on response to a feature detector over scale. A scale-space signature is the vector of measurements at different scales taken at a single point in an image. The system is trained with scale-space signatures from the edge points of a training image. Once trained, the system can determine the degree of "edginess" of points in a new image. A feature-detection framework based on multi-scale analysis and pattern-recognition has several potential advantages over other feature-detection systems. The goal is to create a system that exploits the advantages of a multi-scale, pattern-recognition framework. These advantages are detection of features at different scales (i.e. features of all sizes), robustness to noise, and few or no free parameters. The PASSEF system achieves the stated goals for the detection of step-edge features. The results also show that this technique might be useful in the detection of other features such as crease edges.

This work was conducted by Samuel G. Burgiss while at IRIS lab under the supervision of R. Whitaker (Thesis Chair) and M. A. Abidi. This work was supported by DOE's University Research Program in Robotics under grant DOE-DE-FG02-86NE37968.

 

Papers
  • S. G. Burgiss, E. D. Lester, R. T. Whitaker, and M. A. Abidi, "Scene Segmentation from Vector-Valued Images Using Anisotropic Diffusion", SPIE Conference on Intelligent Robots and Computer Vision XVII: Algorithms, Techniques, and Active vision, Boston, MA, 1998, 3522, pp. 527-538.