Shape Memorization and Recognition of 3D Objects Using a Similarity-Based Aspect-Graph Approach

Abstract

This work proposes an incremental combinational algorithm to generate the prototype of a 3D object using 2D images randomly sampled from a viewing sphere. Similarity-based aspect-graph, which contains a set of aspects and prototypes for these aspects, is employed to represent the database of 3D objects. Furthermore, the proposed algorithm is based on low-level features and similarity measures between the features. In this work, the Fourier descriptor and point-to-point lengths are adopted as features, and three similarity measures, called the 1-norm, 2-norm, and K-L distance, are adopted to extract characteristic views. The effectiveness of the proposed algorithm is demonstrated by experiments with an updating mechanism.

Vision-based Indoor Scene Cognition Using a Spatial Probabilistic Modeling Method

“…Extracting the characteristic views from these 2D views is a kind of approach to obtain a compact set of scene views. In our previous work [16], three kinds of similarity measures, 1-norm, 2-norm and K-L distance, have been applied to extract the characteristic views with the proposed combinational algorithm. In this work, K-L distance is utilized to calculate the similarity between blob-models.…”

“…Then r N ) ( c r N N 2D views are extracted as the set of scene model using a combinational algorithm described in our previous work [16]. If S positions in the environment are selected as the nodes of topological maps, the amount of overall blob models stored in the database can be described as t N , where…”

“…In this work, each scene model is established with 12 Gaussian distributions (N=12) with the 5D feature vectors described in section II. Besides, the characteristic views of the scene model at each position are extracted using the combinational algorithm in our previous work [16] and the amount of the characteristic views at each position are all below 13 after combination ( 11…”

Cortical 3D Face and Object Recognition Using 2D Projections

“…For the recognition of general 3D objects, there are also several approaches which have been explored. Some of these are based on 3D shape retrieval methods (Tangelder & Veltkamp, 2008), on 2D projected images (Su et al, 2006) and on multiresolution signatures (Lam & du Buf, 2011), including approaches that can be applied both to 3D objects and to faces (Passalis et al, 2007).…”