2017-2018 Team: A low power real-time implementation of object detection and tracking on the Zynq 7000 SoC

Team Members: Samuel Rogers, Casey Bui, Nirali Patel, Dishant Patel

https://www.youtube.com/watch?v=3Ve0poDz8rM

Project Overview:

With the expansion of Internet of Things (IoT), computer engineering and science is moving toward the era of IoT-based distributed computing. At the same time, Embedded computer vision is considered one top-tier, fast-growing area. Embedded vision refers to the deployment of visual capabilities to embedded systems for a better understanding of 2D/3D visual scenes. By augmenting the IoT devices with vision processing capabilities many new opportunities and interesting application will be emerged which can elevate the impact of technology in our modern society to the next level.

The aim of this project is to create a distributed IoT system with capabilities to run distributed vision processing across multiple IoT devices. The students will create a network of embedded devices for basic object detection and tracking across multiple cameras. The students will work with multiple boards including Nvidia Jetson TX1/2 and Xilinx Zynq platforms as the IoT devices with the capability of running embedded vision applications. Each board will run locally a simple object tracking algorithm based on the OpenCV open source      vision processing library.

Learning opportunities in this project are many! Overall, accepted student candidates will have a chance to work with Nvidia Jetson TX1/2 and Xilinx Zynq platform and learn embedded Linux, OpenCV embedded vison library and basic distributed processing principles in a very practical way. The results of this project can be magnificent used for a diverse set of applications required distributed vision processing in a large geographical area. Application examples are video surveillance, smart manufacturing and smart and connected communities.

Initial Project Requirements:

The students will work with Nvidia Jetson TX1/2 and Xilinx Zynq boards, embedded cameras, and wireless routers and modems. The equipment’s would be available in my research lab. The students also can use the space available in my research lab to conduct their research. Also, students will able to use the lab servers and computers for development and simulation.

Expected Deliverables/Results:

The students will deliver a implementation of a distributed vision processing over multiple IoT devices using OpenCV object detection and tracking libraries. This includes, implementation of necessary distributed memory storage and communication and processing elements to realize distributed vision processing across heterogeneous IoT devices including Xilinx Zynq and Nvidia Jetson TX1/2 boards.

Disposition of Deliverables at the End of the Project:

A prototyped model of the proposed IoT system with multiple embedded vision devices (Nvidia Jetson TX1/2, Xilinx Zynq) with capability of distributed object detection and tracking will be demonstrated.