Robot Vision Simulation

Robot vision system can get two-dimensional image for the environment by visualsensor, which will be analyses and explained by visual processor, and then be convertedto symbols, through which the robot can recognize object, determine and determine itsposition.This paper summarily describes the analysis and algorithm of the robot visionsystem. Then through the introduction of image semantic model, image semanticextraction and description, this paper discusses some existing problems and certainsolution of semantic-based image retrieval. Focusing on the image semanticclassification and semantic-based retrieval, this paper conducts a study of design andimplement of semantic-based image retrieval system. This paper achieves theenvironmental image acquisition in MSRS simulation platform, and use MATLAB forwavelet image processing and feature extraction value. Finally, it use semantic-basedimage retrieval system to achieve the simulation of robot vision. In this study, theimage retrieval system is realized, it can receive the query model by both keyword andexample image. In the keyword query model, relevance feedback algorithm based onSVM incremental leaning is brought forward. Considering both the relevant results andthe irrelevant ones，SVM classifier is re-trained，the semantic rules are modified，andthe retrieval performance is improved．...