December 2014
Beginner to intermediate
154 pages
4h 7m
English
Feature extraction is an important task in any multimedia retrieval task. Feature extraction has been investigated extensively in recent years. View representation has been improved based on well-defined image feature extraction techniques, which have attracted significant research efforts for decades. Distinct from the features for 2-D images, some shape-based view features are more suitable for 3-D object representation. Popular and effective features, such as moments, Fourier transforms, and the bag-of-words descriptor, are briefly introduced in this chapter. A critical issue is how to combine these views, which may influence the 3-D object description differently. We introduce how we determine weights ...