Tracking with pointing gesture recognition for human-robot interaction

Abstract

In the field of human-robot interaction, it is the key problem for the robot to know the information of the interactive partner. Without obtaining these information the robot couldn't realize what is the partner means. Currently most of the interactive robots have vision sensors to catch human face, and then perform some reactions to the interactive partner. But it could only access under strict conditions because when the partner get out of the sight, the robot couldn't know where to find him. So it is important for the robot to find other information indicated by partner. Some of the interactive robot has audition sensors to localize the sound source. The robot could recognize the location of the sound source produced by partner and try to face him. But the audition system need more computation and hardware resources, so it's hard to be embed in a existing systems. Therefore, we try to find another solution. In this paper, we presented a tracking system with pointing gesture recognition for human-robot interaction. By showing fingers we can provide information for the robot to know that there is another person who wants to interact with it. Finally the whole system will implement at our humanoid robot head.

Hand gesture recognition for Human-Robot Interaction for service robot

“…There are numerous researches for the hand sign, hand gesture and shape recognition by thinning method, contour & curvature, and convex hull based on outstretched hand [19]. In this paper, we choose the distance transform method plus polygonal approximation to obtain the hand's skeleton and then recognize the hand gesture.…”

A High-Security and Smart Interaction System Based on Hand Gesture Recognition for Internet of Things

“…When human interacts with an object by pointing gesture, the pointing direction is estimated by two points on the camera perspective line. According to the different input cameras, pointing gesture recognition technologies can be classified into 3D methods [11][12][13] and 2D methods [14][15][16]. 3D methods rely on specific input devices, such as stereo camera or Kinect, while 2D methods use less expensive and more easily available cameras, with reduced computation cost for 2D information.…”

“…The second challenge is to estimate the pointing position determined by intersection between the interaction plane and the pointing vector due to lack of depth information. In order to solve this problem, some interaction systems based on 2D pointing gesture methods only utilize information of pointing direction instead of pointing position [14,15], and some other systems require users to make coordinate calibration before operating [16]. In this paper, we propose an edge repair-based hand subpart segmentation algorithm, which accurately and effectively segments the palm and finger regions from the background by using 2D information.…”

Chip-based adaptive skin color detection using trajectory constraints on hue

“…In past years, skin color detection has been regarded as an essential component in human-robot interaction systems for analyzing human intentions (Kim et al 2006;Luo et al 2011). This methodology had become popular because it requires a lower computing effort than other image processing approaches even though skin color analysis is a fundamental area of pattern recognition.…”