Low-cost, High-capability, Embedded Systems for Education and Research: A Toolbox for the Microsoft Kinect

Joshua R. Fabian, Tyler Young, James C. Peyton Jones, Garrett M. Clayton · 2020

Abstract Low‐cost,High‐Capability,EmbeddedSystemsforEducationand Research:ToolboxesforthePandaBoardandMicrosoftKinect Authors: left blank Affiliation: left blankIncreasingly powerful, yet low cost computing and sensing devices are now available for use instudent designs and embedded mechatronic systems. The PandaBoard, for example, is a lowcost board running full Ubuntu Linux on a dual core ARM Cortex processor. It supports HDvideo, and has onboard BlueTooth, WiFi, USB and serial communications. The MicrosoftKinect, though initially developed as a gaming device, provides rich sensing possibilities, withcamera and depth images, again at remarkably low cost. However, the sophistication of thesedevices often requires a high degree of programming ability in order to exploit their evidentcapabilities. This paper describes progress on a National Science Foundation and MathWorkssponsored project aimed at making these devices more accessible to student users through theuse of Automatic Code Generation techniques. Specifically, the paper describes several newtoolboxes that have been developed which allow students to perform their designs from withinthe Matlab / Simulink environment, and then to implement these designs directly on aPandaBoard and/or Kinect system. Students develop their designs using interconnectedSimulink blocks and subsystems, and the build process automatically cross-compiles anddownloads the model to the target for execution. External mode capability can be used tomonitor the target hardware as it executes in real time, enabling the user to tune modelparameters and log data while their application is running. An example showing how thePandaBoard and Kinect can be integrated as a single system is shown as an example.

Read the paper · More papers on PaperTik