Hidden Markov Model Based Visual Perception Filtering in Robotic Soccer

International Journal of Advanced Robotic Systems

关键词: Robotic Soccer;    Perception;    Robot cognition;    Statistical Inference;    Localization;   

DOI  :  10.5772/7253

学科分类：自动化工程

来源: InTech

【 摘 要 】

Autonomous robots can initiate their mission plans only after gathering sufficient information about the environment. Therefore reliable perception information plays a major role in the overall success of an autonomous robot. The Hidden Markov Model based post-perception filtering module proposed in this paper aims to identify and remove spurious perception information in a given perception sequence using the generic meta-pose definition. This method allows representing uncertainty in more abstract terms compared to the common physical representations. Our experiments with the four legged AIBO robot indicated that the proposed module improved perception and localization performance significantly.

【 授权许可】

CC BY