Towards robust place recognition for robot localization

Abstract

Abstract-Localization and context interpretation are two key competences for mobile robot systems. Visual place recognition, as opposed to purely geometrical models, holds promise of higher flexibility and association of semantics to the model. Ideally, a place recognition algorithm should be robust to dynamic changes and it should perform consistently when recognizing a room (for instance a corridor) in different geographical locations. Also, it should be able to categorize places, a crucial capability for transfer of knowledge and continuous learning. In order to test the suitability of visual recognition algorithms for these tasks, this paper presents a new database, acquired in three different labs across Europe. It contains image sequences of several rooms under dynamic changes, acquired at the same time with a perspective and omnidirectional camera, mounted on a socket. We assess this new database with an appearancebased algorithm that combines local features with support vector machines through an ad-hoc kernel. Results show the effectiveness of the approach and the value of the database.

Object Templates for Visual Place Categorization

“…Pronobis et al also published several datasets in order to set up standard benchmark [4,12]. In [13], Ullah et al used a SVM-based method to address place recognition and categorization on the dataset COsy Localization Database (COLD) [12].…”

“…We only tested 4 categories that are available for all 3 labs: Printer Area (PA), Corridor (CR), 2-Person Office (2PO) and Bathroom (BR). [13] tested their place recognition and categorization system on COLD, their categorization testing method is similar with ours, so we compare our result with them in Table 3. Note that since they only provide histogram of their result with no exact numbers, their accuracy is estimated from the histogram.…”

“…Table 3. Comparing our system with [13]. [13]'s result is from cloudy condition, given that we also use sequences from cloudy conditions.…”

Indoor Scene Classification Using Combined 3D and Gist Features

“…Known rooms can be recognized by invariant features in 360 • laser scans [5], by comparing current 2D features to saved views [6], and by recognizing specific objects and their configurations [7]. Burgard's group has examined a huge bunch of different approaches to the more general place categorization problem ranging from the detection of concepts like "room", "hall", and "corridor" by simple geometric features defined on a 2D laser scan [8] to subconcepts like "kitchen" [9] using ontologies encoding the relationship between objects and the subconcepts [10].…”

“…They partition an image into increasingly fine sub-regions, the so-called spatial pyramid, and represent these regions through SIFT descriptors. In [6] the application of such local features for place categorization on a mobile robot in a real environment is tested with moderate success.…”