Extraction of Haar integral features on onmidirectional images: Application to local and global localization

In this paper, we present a new method for producing omnidirectional image signatures that are purposed to localize a mobile robot in an office environment. To solve the problem of perceptual aliasing common to the image based recognition approaches, we choose to build signatures that greatly vary between rooms and slowly vary inside a given room. We suggest an averaging technique based on Haar integral invariance. It takes into account the movements the robot can do in a room and the omni image transformations thus produced. The variability of the built signatures is adjusted (total or partial Haar invariance) according to defined subsets of the group transformation. The experimental results prove to get significantly interesting results for place recognition and robot localization with variable accuracy: From global rough localization to local precise one.