ARM TechCon 2015 Schedule

All SessionsSpeakersMy Schedule
View Sessions As:
  • De-Mystifying Automotive ADAS Collision Avoidance systems with Programmable SoCs

    Dan Isaacs  |  Director, Xilinx
    Location:  Ballroom F
    Format: 50-minute Technical Session
    Track: Automotive/Embedded Vision
    Pass Type: Conference Pass - Get your pass now!
    Vault Recording: TBD
    Audience Level: All

    From automated 3D-image stitching for surround-view, to optical-flow, image processing in ADAS systems has become increasingly processing and bandwidth driven. Managing multiple cameras' input real-time with megapixel resolution requires high performance processing and low latency.

    Combining real-time and application processing and flexible programmable logic for high performance parallel processing, multi-core programmable SoCs provide an ideal platform for implementing vision-based collision avoidance systems.

    This paper presents an analysis of critical requirements to implement these embedded vision systems and how the latest multi-core processor programmable SoCs meet these needs through functional partitioning between real time and application processing and programmable logic.


    Attendees will learn about key requirements driving next generation Automotive driver assistance applications

    Learn latest technology for developing ADAS systems using ARM-based heterogeneous multicore programmable SoCs including processing functionality prioritizations between application and realtime processing.

    Learn latest software development environment for optimizing functional partitioning between processing and programmable logic algorithmic acceleration.