FEATURE EXTRACTION FROM VRML MODELS FOR VIEW-BASED OBJECT RECOGNITION

Thomas E Kruger, Jochen Wickel, Pablo Velarde-Alvarado, Karl–Friedrich Kraiss · 2003

View-based object recognition offers approaches to recognize three-dimensional objects from arbitrary two-dimensional views. In this paper we suggest different simple feature extraction methods based on three-dimensional object models. These are suited for both generating symbolic object descriptions and comparison with shape features extracted from images. We present a concept for improving view-based recognitions and provide first experimental results. 1.

Read the paper · More papers on PaperTik