MACC'97 Poster session
An image understanding system for real world images which has an ability to recognize various kinds of images is proposed. We propose a multi-agent architecture to integrate object recognition modules for many kinds of target objects. In our method, recognized results by several different agents are fused based not only on the evaluations by each modules themselves but also on relations of object locations, sizes and so on. This is carried out autonomously between the agents concerned, and the most reliable result is selected after the arbitration between them. We implemented an experimental system on a parallel computer, and achieved recognition for both indoor and outdoor images.
image understanding system, multi-agent, object recognition, scene interpretation
PS
file(+gzip) (in Japanese)