Feature Silhouette Extraction from Photographs for Virtual Human Modeling

In this paper, we present an easy, efficient and practical algorithm, which extracts the feature silhouette from a photograph for virtual human modeling. Our segmentation algorithm is derived from the MumfordShah segmentation technology and the level set formulation. We accelerate the silhouette extraction process using the multi-pyramid level method. After that, a feature extraction algorithm is introduced to determine the feature points on the human contours. At the end of the paper, some results of the virtual human modeling are shown to demonstrate the functionality. Compared with other approaches, our silhouette extraction methods and feature extraction are automatic, and the speed of extraction is fast. KEY WORDS: Active contour, Segmentation, Feature extraction, 3-D Object Extraction.

• Type: Conference
• Year: 2001
• Where: VIIP
• Authors: Yu Wang 0010, Charlie C. L. Wang, Matthew Ming-Fai Yuen
• Keywords: Feature Extraction, Image Processing, Silhouette Extraction, VIIP 2001, Virtual Human Modeling