Kinect Human Follower
No Thumbnail Available
Date
2013
Journal Title
Journal ISSN
Volume Title
Publisher
Uva Wellassa University of Sri Lanka
Abstract
In near future people will need their own personal robot for day to day work. With this trend
education,
entertainment and security, especially in defense, the market for robots will be enormously high
(Saccoccio & Taleb, 2013). The range of the camera in Microsoft's Kinect, intended for the Xbox
360 gaming console, offers a powerful alternative to many standard sensors used in robotics for
gathering spatial information about a robot’s surroundings (Schwab, 2011). The recently released
Kinect is the first commercially available product which provides the depth data of its resolution
and accuracy reach of many robotics projects. This project aims to design and construct a mobile
autonomous robot which can navigate independently taking gesture commands from Kinect
sensor in a very cost effective way in order to be used in home application (MikkelViager, 2011).
Methodology
The project is divided into two parts: hardware implementation and software implementation.
The platform provides the robot’s mobility and serves as the base of the robot (Schwab, 2011).
All of the other components are mounted on the base. The steering consists of two links. One
steering link consists of a medium caster wheel and other consisting with two wheels driven by
12 Volt two DC motors. The circuit design consists of microcontrollers circuit with max 232 for
serial communication. Motor controlling circuit was constructed using H-Bridge. The system
software can be broken down into several components that approximately correspond to the
hardware components. OpenNI, NITE from PrimeSense were downloaded and installed.Visual
Studio 2008 was installed to further implementation of Kinect sensor.
Results and Discussion
Overall goal of this project was to identify the commands and Kinect’s ability to provide
sufficient sensory information to perform basic navigation.
The first program was tested to identify the ability of the device. Therefore invincible pen was
created using hand. Procedure was completed successfully to draw lines using invincible pen.The
system was tested to identify the depth image. Kinect could able to identify the depth image.
Closer objects are dark in color and objects that are further away are lighter in color of the depth
image. Figure 1 depicts the Kinect depth data represented as an image.Also of importance, there
are noticeable black regions in this image that Kinect can’t identify.
To take the gesture command Prime sense NITE should be linked with openNI.Visual studio
2008 unable to link Prime sense NITE with openNI. Therefore the Kinect unable to perform
gesture commands.
The motors successfully operated for a long time and delivered a better torque to the wheels in
order to carry higher weights. Motor controlling of the robot was successfully actuated. The
circuit which was prepared for serial communication was successfully operated. It could able to
communicate with PC and Microcontroller.
Description
Keywords
Science and Technology, Technology, Computer Science, Robotics, Human Robotics