Visual navigation method for indoor mobile robot based on extended BoW model

• Published in: CAAI Transactions on Intelligence Technology Vol. 2; no. 4; pp. 142 - 147
• Main Authors: Li, Xianghui, Li, Xinde, Khyam, Mohammad Omar, Luo, Chaomin, Tan, Yingzi
• Format: Journal Article
• Language: English
• Published: Beijing The Institution of Engineering and Technology 01.12.2017; John Wiley & Sons, Inc; Wiley
• Subjects: (SIFT)-detection algorithm; Algorithms; angles; B6135 Optical, image and video signal processing; B6135E Image recognition; C3390C Mobile robots; C5260B Computer vision and image processing techniques; dynamic indoor environments; extended bag; extended BoW model; feature extraction; feature points; Feature recognition; feature vectors; general object recognition; graphic processing unit acceleration technology; Graphics processing units; human-robot interaction; human–robot interaction method; Indoor environments; indoor mobile robot; Libraries; mobile robots; Navigation; navigation technology; Object recognition; original BoW model; redundant image information; relative distances; robot vision; Robots; scale-invariant feature; Special Issue: Recent Advances in Intelligent Robotic Systems; statistical information; support vector machine classifier; Support vector machines; transforms; unknown environments; visual navigation method; words model; object recognition; dynamic indoor environments; indoor mobile robot; angles; unknown environments; scale-invariant feature; support vector machine classifier; transforms; mobile robots; extended bag; human–robot interaction method; statistical information; visual navigation method; feature extraction; navigation technology; original BoW model; feature points; graphic processing unit acceleration technology; human-robot interaction; support vector machines; words model; robot vision; general object recognition; extended BoW model; (SIFT)-detection algorithm; feature vectors; relative distances; redundant image information
• ISSN: 2468-2322; 2468-2322
• DOI: 10.1049/trit.2017.0020

This article proposes a new navigation method for mobile robots based on an extended bag of words (BoW) model for general object recognition in indoor environments. The scale-invariant feature transform (SIFT)-detection algorithm with the graphic processing unit (GPU) acceleration technology is used to describe feature vectors in this model. First, in order to add some redundant image information, statistical information of the spatial relationships of all the feature points in an image, i.e. relative distances and angles, is used to extend the feature vectors in the original BoW model. Then, the support vector machine (SVM) classifier is used to classify objects. Also, in order to navigate conveniently in unknown and dynamic indoor environments, a type of human–robot interaction method based on a hand-drawn semantic map is considered. The experimental results show that this new navigation technology for indoor mobile robots is very robust and highly effective.