Next-generation automotive image processing with ARM Mali-C71

Similar documents
Next-generation automotive image processing with ARM Mali-C71

Functional safety for semiconductor IP

Automotive In-cabin Sensing Solutions. Nicolas Roux September 19th, 2018

Enabling Mobile Virtual Reality ARM 助力移动 VR 产业腾飞

Sony Releases the Industry's Highest Resolution Effective Megapixel Stacked CMOS Image Sensor for Automotive Cameras

Transforming Industries with Enlighten

The Denali-MC HDR ISP Backgrounder

Image Processing Architectures (and their future requirements)

Vision with Precision Webinar Series Augmented & Virtual Reality Aaron Behman, Xilinx Mark Beccue, Tractica. Copyright 2016 Xilinx

A Winning Combination

Radar and Functional Safety technology for advanced driving assistance

The Building Blocks of Autonomous Control. Phil Magney, Founder & Principal Advisor July 2016

Image Processing Architectures (and their future requirements)

Combining ROS and AI for fail-operational automated driving

FLEA 3 GigE Vision FLIR IMAGING PERFORMANCE SPECIFICATION. Version 1.1 Revised 1/27/2017

Automated Driving Systems with Model-Based Design for ISO 26262:2018 and SOTIF

FLASH LiDAR KEY BENEFITS

SEE MORE, SMARTER. We design the most advanced vision systems to bring humanity to any device.

EM Simulation of Automotive Radar Mounted in Vehicle Bumper

Digital Engines for Smart and Connected Cars By Bob O Donnell, TECHnalysis Research Chief Analyst

The IQ3 100MP Trichromatic. The science of color

Movie 3. Basic Camera Raw workflow

IHV means Independent Hardware Vendor. Example is Qualcomm Technologies Inc. that makes Snapdragon processors. OEM means Original Equipment

Table 1: Example Implementation Statistics for Xilinx FPGAs. Fmax (MHz) LUT FF IOB RAMB36 RAMB18 DSP48

A NEW NEUROMORPHIC STRATEGY FOR THE FUTURE OF VISION FOR MACHINES June Xavier Lagorce Head of Computer Vision & Systems

23270: AUGMENTED REALITY FOR NAVIGATION AND INFORMATIONAL ADAS. Sergii Bykov Technical Lead Machine Learning 12 Oct 2017

Take Mobile Imaging to the Next Level

BLACKFLY USB3 Vision FLIR IMAGING PERFORMANCE SPECIFICATION. Version 5.1 Revised 1/26/2017

VSI Labs The Build Up of Automated Driving

Physics Based Sensor simulation

Evaluating Functional Safety in Automotive Image Sensors

Virtual Testing of Autonomous Vehicles

Improvements in Functional Safety of Automotive IP through ISO 26262:2018 Part 11

[Overview of the Consolidated Financial Results]

Software ISP Application Note

CONVERTING AND EDITING RAW IMAGES

Converting and editing raw images

Faurecia : Smart Life on board An innovative company

HDR with Smart Objects

Autonomous Automation: How do we get to a Million Miles of testing?

Dealing with the Complexities of Camera ISP Tuning

Striker II. Performance without compromise

EF-45 Iris Recognition System

About 3D perception. Experience & Innovation: Powered by People

High dynamic range imaging and tonemapping

Dr George Gillespie. CEO HORIBA MIRA Ltd. Sponsors

32 Float v3 Quick Start Guide. AUTHORED BY ANTHONY HERNANDEZ (415)

High Performance Imaging Using Large Camera Arrays

It should also be noted that with modern cameras users can choose for either

Camera Image Processing Pipeline: Part II

An Introduction to Histograms in Photography

Mission Space. Value-based use of augmented reality in support of critical contextual environments

This document is a preview generated by EVS

ADAS Development using Advanced Real-Time All-in-the-Loop Simulators. Roberto De Vecchi VI-grade Enrico Busto - AddFor

Education and Training

The GRAIL project: Galileo Localisation for the European Train Control System

A Roadmap for Connected & Autonomous Vehicles. David Skipp Ford Motor Company

TECHNOLOGY DEVELOPMENT AREAS IN AAWA

OVERVIEW WHERE TO FIND THE SETTINGS. CION Technical Notes #1 Exposure Index, Gamma and In-Camera Color Correction Comparison

Adaptive Vision Leveraging Digital Retinas: Extracting Meaningful Segments

PHOTO 11: INTRODUCTION TO DIGITAL IMAGING

ALMALENCE SUPER SENSOR. A software component with an effect of increasing the pixel size and number of pixels in the sensor

Responsible Data Use Assessment for Public Realm Sensing Pilot with Numina. Overview of the Pilot:

A Matter of Trust: white paper. How Smart Design Can Accelerate Automated Vehicle Adoption. Authors Jack Weast Matt Yurdana Adam Jordan

Photomatix Light 1.0 User Manual

Camera Image Processing Pipeline: Part II

Visualization in automotive product development workflow

Volkswagen Group: Leveraging VIRES VTD to Design a Cooperative Driver Assistance System

Camera Requirements For Precision Agriculture

William Milam Ford Motor Co

PRESS INFORMATION. Nikon Introduces Next Generation Image-Editing Software. Capture NX2: more powerful and easier to use

Deliverable D1.6 Initial System Specifications Executive Summary

Industrial Keynotes. 06/09/2018 Juan-Les-Pins

FAIL OPERATIONAL E/E SYSTEM CONCEPT FOR FUTURE APPLICATION IN ADAS AND AUTONOMOUS DRIVING

VGA CMOS Image Sensor BF3905CS

Neural Networks The New Moore s Law

Development of a 24 GHz Band Peripheral Monitoring Radar

Retina 400 THE SIMPLE AND FLEXIBLE SOLUTION FOR HIGH DEFINITION RETINAL IMAGERY NON-MYDRIATIC RETINAL CAMERA C/D AUTO 60 H X 45 V LED. 2,2 mm.

Evaluation of Connected Vehicle Technology for Concept Proposal Using V2X Testbed

HIGH DYNAMIC RANGE VERSUS STANDARD DYNAMIC RANGE COMPRESSION EFFICIENCY

White paper on CAR28T millimeter wave radar

Advanced Near-Realistic 3D Lighting: Capturing, Creating, and Using Image-Based Lighting

Photoshop Elements for Genealogists Part 1

Photographic Exposure Colin Legg

Global Image Sensor Market with Focus on Automotive CMOS Sensors: Industry Analysis & Outlook ( )

The Xbox One System on a Chip and Kinect Sensor

This histogram represents the +½ stop exposure from the bracket illustrated on the first page.

Electronic Costing & Technology Experts

76-81GHz MMIC transceiver (4 RX / 3 TX) for automotive radar applications. Table 1. Device summary. Order code Package Packing

Use of Photogrammetry for Sensor Location and Orientation

Camera Requirements For Precision Agriculture

DEEP LEARNING A NEW COMPUTING MODEL. Sundara R Nagalingam Head Deep Learning Practice

Teton Photography Group

CORRECTED VISION. Here be underscores THE ROLE OF CAMERA AND LENS PARAMETERS IN REAL-WORLD MEASUREMENT

Multi-sensor Panoramic Network Camera

Transportation Informatics Group, ALPEN-ADRIA University of Klagenfurt. Transportation Informatics Group University of Klagenfurt 3/10/2009 1

Basic Camera Craft. Roy Killen, GMAPS, EFIAP, MPSA. (c) 2016 Roy Killen Basic Camera Craft, Page 1

AUTOCOM VEHICLE INSPECTION SERVICES

PHOTOGRAPHY: MINI-SYMPOSIUM

LEARNING FROM THE AVIATION INDUSTRY

Transcription:

Next-generation automotive image processing with ARM Mali-C71 Chris Turner Director, Advanced Technology Marketing CPU Group, ARM ARM Tech Forum Korea June 28 th 2017

Pioneers in imaging and vision signal processing 2

What is an ISP? The Image Signal Processor (ISP) processes raw pixel data into great-looking images for display, encoding and further processing by computer vision algorithms 3

Automotive needs: viewability and reliability 4

Automotive needs: viewability and reliability 5

Automotive needs: viewability and reliability Pedestrian protection system Driver drowsiness detection Night vision Low latency Reliability Resistant to random and systematic faults Imagery feeds into computer vision engine Viewability Dynamic range Low-light and low-noise performance Imagery close to human eyes and at times beyond 6

Automotive ADAS growth Camera volume rising very rapidly, around 20% CAGR By 2023 (expected average): Mid-range (high volume) cars: three cameras Luxury sedans: <ten cameras 7

Advanced vision requires centralised processing ISP per camera sensor Less flexible Higher BOM costs Centralised sensor fusion SoC Centralised processing Flexible architecture Sensor fusion SoC 8

And there are standards!! A number of functional safety standards exist: ISO 26262 Road vehicles IEC 61508 Electrical, electronic programmable systems DO 254 Electronics that fly e.g. in airplanes, helicopters 9

ISO 26262-2011 road vehicles functional safety Relates to electrical and electronic systems used in automobiles Qualitatively assesses risk of hazardous operational situations Aims to avoid or control systematic failures and detect or control random hardware failures, or mitigate their effects 10

ISO 26262: goals Provides an automotive safety lifecycle and supports tailoring the necessary activities during these lifecycle phases Covers functional safety aspects of the entire development process Provides an automotive-specific risk-based approach for determining risk classes (Automotive Safety Integrity Levels, ASILs) Provides requirements for validation and confirmation measures to ensure a sufficient and acceptable level of safety is being achieved 11

High-performance ISP for automotive: ARM Mali -C71 Ultra WDR Multi-camera input Designed for safety Ultra-wide dynamic range 24 stops Every detail captured Simultaneous support for human display and computer vision Virtual ISPs: <4 real-time camera inputs Pleasing and natural for human vision; predictable and reliable for computer vision Every pixel reliable Real-time safety for ADAS applications Enabling system-level ASIL D and SIL3 compliance 300+ fault detection circuits, built-in self-test (BIST), CRC on data paths, every pixel tagged 12

Mali-C71 block diagram Image sensor sources Colour/data plane outputs 13

Enabling system level ASIL D compliance Random faults Systematic faults Safety features 300+ dedicated fault detection circuits Built in continuous self-test Fault interrupt controller Pixel consistency image plane tagging reliability of every pixel Detection of sensor and hardware faults CRC checks to all data paths including memories and configurations Processes Developed within robust requirements, tracing and validation framework Safety manual Failure modes and effects analysis (FMEA) Development interface report Providing support for system level: SIL 3 / IEC 61508 ASIL D / ISO 26262 With a standards-agnostic approach 14

ARM Mali-C71 Image Signal Processor Designed to enable system-level certification Safety Element out of Context (SEooC) ISP has no direct control authority System integrating ARM s ISP needs to meet the requirements ISO26262 (ASIL D) IEC 61508 SIL3 Standard product Integrated into system (SoC) by our customers Specific use cases unknown to ARM Has to detect faults, perform self-tests Has a responsibility to provide timely, accurate and reliable information to the system Required for the system to be safe Road vehicles functional safety This flows down to the components of the system 15

Supporting output for human and machine vision Human display Computer vision 16

Ultra-wide dynamic range 2^24 2^8 Highlights clipped Mid-tones not revealed 0 2^8 Shadows details lost Preserved highlights As perceived by the human eye Enhanced mid-tones 0 0 Revealed shadows 17

Dynamic range management: iridix More than just the world s best local tonemapping engine Precise model of human retina contrast adaption Precisely simulates illumination by white light Basis of Nikon D-Lighting (2005) Shipped in over 2 billion devices iridix lets cameras see like the human eye 18

Dynamic range iridix engine: pixel-by-pixel tone mapping Contrast in highlights enhanced Contrast in midtones preserved iridix calculates a unique tone curve for each pixel of each frame Contrast in shadows enhanced 19

Management of dynamic range 20

Management of dynamic range 21

A complete, high-performing solution Performance Software Tuning 1.2 Giga-pixels/sec of processing performance Supports up to 4 real-time camera inputs and 16 camera streams with a single pipeline Control the ISP, sensor, auto-white balance and auto-exposure Roadmap to complete automotive software designed for ASIL compliance Full set of tools provided Ecosystem support for tuning and bringing up specific use-cases and sensors 22

Mali-C71: the next step in automotive vision First automotive ISP with ultra-wide dynamic range Supports both human display and computer vision Enables safer, more comfortable driving 23

The trademarks featured in this presentation are registered and/or unregistered trademarks of ARM Limited (or its subsidiaries) in the US and/or elsewhere. All rights reserved. All other marks featured may be trademarks of their respective owners. Copyright 2017 ARM Limited