Over-the-Shoulder Tutorial: Accelerating Time to an Image Processing Prototype


Date:  Wednesday, September 16, 2020

Start Time: 11:00 am

End Time: 11:45 am

During this walkthrough session, attendees will learn how they can go from an initial concept/architecture of an embedded vision design to demonstrating application and algorithm feasibility using off-the-shelf prototyping systems from Avnet and Xilinx.

The session will demonstrate key steps used to create a viable prototype using the Ultra96-V2 and Xilinx solutions:
∙ Creation of an image processing pipeline using Vivado
∙ Testing the image processing pipeline on the Ultra96 using an open-source Python-based framework and OpenCV with the Dual MIPI Board and MIPI cameras running algorithms on the processor cores
∙ Acceleration of the vision algorithm using Vitis and OpenCL to create a final application

Track

Session Speakers

  • Over my career, having worked in a wide variety of fields, from Automotive Safety Test equipment, VoIP, Telco systems, Thermal Imaging Camera, and Satellite / Terrestrial Radio Communications, I have been fortunate enough to have a wide variety of experience allowing me to be in a unique position having seen and worked with everything from gate logic, to FPGA and ASIC design.  The most fun I've had was working with FPGAs where you do not have the pressure of 'cutting an ASIC,' yet you get the fun of designing your OWN chip!  Now, being in Avnet's Products & Emerging Technologies group, I currently work with all of our boards and have a major focus on the UltraZed family.

  • Kevin Keryk is a Software Engineering Manager and Machine Learning Specialist with Avnet. He is a Computer Engineer by degree, Technical Marketing Manager by day, and an embedded software programming addict from an impressionable age which has turned into a mostly-healthy fascination with FPGA and other hardware craft over the past 10 years at Avnet.

  • Adam Taylor is a chartered engineer and fellow of the Institute of Engineering and Technology. Over his multi-decade career, he has had experience within the public and private sector, developing FPGA-based solutions for a range of applications including RADAR, nuclear reactors, satellites, cryptography and image processing. Adam is also a Visiting Professor of embedded systems at the University of Lincoln. Education is his deepest passion and he has delivered thousands of hours' worth of training to corporate clients and casual hardware enthusiasts alike.


Share

Over-the-Shoulder Tutorial: Accelerating Time to an Image Processing Prototype