The Selfie Robot: An Integrated Cloud Powered Framework

 

Introduction

Two of our interns in the DPI Community Evangelism group, Isabel Zhang and Milly Xun, have created an interesting IoT/Robotics project over the past two months.

 

 

The project features a robot arm with a mounted webcam that detects and tracks objects, while moving the arm to keep the object centered in frame. The robot arm can also snap your picture when you smile, uploading the image to Amazon Rekognition (a deep learning image processing service) for image processing. It is able to pull out information about the person's gender, location, and many other useful tidbits. All local processing is done on an Intel i7 NUC.

 

 

This project integrates a whole host of technologies into a framework that can be used for any number of applications.

 

The Intel NUC is running Ubuntu Xenial* 16.04 <releases.ubuntu.com/16.04/>

The real-time computer vision in done with OpenCV* <http://opencv.org/>

The robotic arm control as well as cross node communication is done with ROS Kinetic* <http://www.ros.org/>

The full-stack development for the front end webpage is done with MEAN Stack <http://meanjs.org/>

The cloud service integration is done with AWS* IoT and AWS* Rekognition* <https://aws.amazon.com/rekognition/>

 

Our hardware list:

Computer: Intel i7 NUC <https://www.intel.com/content/www/us/en/products/boards-kits/nuc.html>

Robot Arm: PhantomX* Reactor Robotic Arm with Wrist <http://www.trossenrobotics.com/p/phantomx-ax-12-reactor-robot-arm.aspx>

Webcam: Logitech* c270 <http://support.logitech.com/en_us/product/hd-webcam-c270>


 

Make sure to check in later for updates with code samples and explanations!

 

*Test Change*

 

For more information:

Learn about IoT

Learn about IoT

For more complete information about compiler optimizations, see our Optimization Notice.