Seminář: A System for Appearance-Based Probabilistic 3D Object Recognition and Its Applications

Within the scope of this presentation a system for appearance-based statistical classification and localization of 3D objects in 2D digital images will be presented. In contrast to shape-based approaches, appearance-based methods do not use any segmentation steps to extract object features. The objects are described by 2D local feature vectors computed directly from image pixel values using the wavelet transformation. Both gray level and color images can be used for feature extraction. Finally, the object features are statistically modeled with the normal distribution and stored in the object models as density functions. In the recognition phase the system classifies and localizes objects in scenes with real heterogeneous background, whereas the number of objects in a scene is unknown. First, feature vectors are calculated in the scene with the same method as in the training. Second, a maximization algorithm compares the learned density functions with the extracted feature vectors and yields classes and poses of objects found in the scene. Experiments made on a real dataset with more than 40000 images prove high robustness of the system in terms of classification and localization rates. Finally, the idea of adapting the system to several real world tasks will be pointed out.