Categorization of Indoor Places Using the Kinect Sensor

Abstract

The categorization of places in indoor environments is an important capability for service robots working and interacting with humans. In this paper we present a method to categorize different areas in indoor environments using a mobile robot equipped with a Kinect camera. Our approach transforms depth and grey scale images taken at each place into histograms of local binary patterns (LBPs) whose dimensionality is further reduced following a uniform criterion. The histograms are then combined into a single feature vector which is categorized using a supervised method. In this work we compare the performance of support vector machines and random forests as supervised classifiers. Finally, we apply our technique to distinguish five different place categories: corridors, laboratories, offices, kitchens, and study rooms. Experimental results show that we can categorize these places with high accuracy using our approach.

Categorization of indoor places by combining local binary pattern histograms of range and reflectance data from laser range finders

“…Our work is close to [20], where a Kinect camera is used to categorize pairs of depth and gray scale images into indoor places using histograms of local binary patterns. However, the working range of the Kinect camera is very short.…”

Towards generalization of semi-supervised place classification over generalized Voronoi graph

“…Using the multi-class SVM on features extracted from real 3D data, Swadzba and Wachsmuth [23] achieved approximately 80% accuracy, while the work of Mozos et al [24] showed accuracies above 92% and we have obtained approximately 97% [25]. It is to be noted that the accuracies mentioned here are achieved by different methods, on different data sets and may address completely different tasks.…”

“…In a similar setup, we implemented both binary and multi-class logistic regression based solutions, and have been able to achieve accuracies above 98% [21,22]. In recent years, SVM as a prominent classifier, has been gaining popularity over other approaches in many applications [23,24].…”

Database for semantic model of the operating environment of mobile service robot

“…В этом направлении ведется разработка специализированных языков для описания среды, алгоритмов построения семантических моделей на основании данных сенсоров, создаются специализированные базы семантической информации, [1], [2], [3]. Однако в настоящее время, несмотря на значительные усилия, предпринимаемые мировыми производителями, на рынке практически нет роботов, которые использовали бы семантическую информацию для построения целенаправленного поведения.…”