Object recognition and localization via pose clustering
Computer Vision, Graphics, and Image Processing
Recognizing 3-D Objects Using Surface Descriptions
IEEE Transactions on Pattern Analysis and Machine Intelligence
SIGMA: A Knowledge-Based Aerial Image Understanding System
SIGMA: A Knowledge-Based Aerial Image Understanding System
Relational matching with dynamic graph structures
ICCV '95 Proceedings of the Fifth International Conference on Computer Vision
Document processing with Bayesian network and agent-based programming
SSIP'05 Proceedings of the 5th WSEAS international conference on Signal, speech and image processing
| Hi-index | 0.00 |
This paper describes a feasibility study of "multi-agent oriented" techniques on a 2-D and 3-D object recognition system. The main aim of the project is to develop an inspection supporting tool that understands objects in both 2-D and 3-D in a unified system. 2-D and 3-D worlds are mapped to each other via agent-like entities each of which holds a conceptualized representation, allowing for a robust inference ability. Agents, each of which has symbolic representation of a part of an object, are hierarchically organized to represent a complete representation of an object. In this paper, object recognition is carried out with two matching methods: (1) the matching between an object model (agent's knowledge) and observed data, and (2) a constraint propagation-like method to achieve overall consistency among agents. The first is carried out with a symbolic Hopfield-type neural network and the second via a hierarchical Winner-Takes-All algorithm.