Overview:
In this presentation, you will learn how to
rapidly accelerate real-time computer vision algorithms and integrate them into
your designs using Xilinx Zynq®7000 All Programmable SoC devices. Step-by-step
examples will show you how to build and integrate custom video accelerators by
migrating functions from the OpenCV library or your own custom video algorithms
into your embedded design using guaranteed-to-synthesize HLS (High-level
Synthesis) video libraries; the Vivado® HLS compiler; and the Vivado IP
Integrator. This design flow allows you to implement real-time embedded
computer-vision algorithms using a seamless combination of high-performance,
low-power, on-chip programmable hardware for high-data-rate (full HD)
pixel-processing tasks and software-programmable ARM processor cores for
frame-based processing tasks that operate at lower data rates.

What attendees will learn:

  • Xilinx Smarter Vision solutions’ for computer vision
  • How to implement OpenCV functions using Zynq-7000 All Programmable SoCs.
  • How to accelerate OpenCV function calls by replacing them with
    HLS-synthesizable function calls that run as much as 700x faster than the same
    functions implemented in software.
  • How to refactor OpenCV functions by encapsulating them in a hardware
    accelerator with standard AXI4 interfaces and easily connect these hardware
    locks to the rest of your system.
  • How to rapidly integrate Vivado HLS created accelerator functions using the
    Vivado IP Integrator in a reference design based on the Zynq-7000 All
    Programmable SoC ZC702 Eval Board.

Presenters:

Stephen Neuendorffer, Principal Engineer, Vivado High-Level
Synthesis, Xilinx

Stephen Neuendorffer is a Principal Engineer and
Product Architect in the High-Level Synthesis group at Xilinx. He has been
working at Xilinx for almost nine years focusing on High-Level Synthesis and
improving the usability of programmable logic devices for software engineers.
Prior to working in the product group he spent seven years in the Xilinx
research labs. Stephen received B.S. degrees in electrical engineering and
computer science from the University of Maryland, College Park, in 1998. He
graduated with University Honors, Departmental Honors in electrical engineering,
and was named the Outstanding Graduate in the Department of Computer Science. He
received the Ph.D. degree from the University of California, Berkeley, in 2003,
after being one of the key architects of Ptolemy II. He has more than 30
publications and 14 patents.

Girish Malipeddi, Video Solutions Marketing Manager,
Xilinx

Girish Malipeddi is a Sr. Product Manager for video solutions
at Xilinx. Girish has been in the FPGA industry for over 13 years and is
responsible for defining programmable logic device based video solutions. Prior
to that, he was doing Product Applications and Marketing at Audience and Memory
design at Alliance Semiconductor. Girish holds M.S.E.E from the University of
Alabama in Huntsville and a B.S.E.E from Osmania University, Hyderabad, India.
He has published articles on Programmable logic devices for Video applications.