Beginning of this page
Jump to main content

Please note that JavaScript and style sheet are used in this website,
Due to unadaptability of the style sheet with the browser used in your computer, pages may not look as original.
Even in such a case, however, the contents can be used safely.


Worldwide > Japan日本語

What is IMAPCAR?


Sample Pre-crash Safety Feature Using IMAPCAR Video Recognition

Image Input (Camera) → IMAPCAR Image Processing → Warning the driver Activating automatic braking system



Our Approach

Utilizing video recognition as a core improvement to vehicle safety will require highly reliable mass produced products, and the flexibility to make technical innovations rapidly. Until now, however, the only way to provide the real-time processing required for video recognition was with large, dedicated hardware.

IMAPCAR uses software for the entire processor base. This makes it easy to develop, modify, and upgrade. Today, nearly all cars are equipped with airbags, which has greatly reduced the number of fatalities. The number of traffic accidents, however, actually continues to increase. In the future, we can expect the number of traffic accidents to fall through the use of cars that avoid collisions via preventive safety systems using less technology. IMAPCAR is a major step toward making this a reality, delivering the video recognition technology that has until now been a major stumbling block.


A highly parallel processor performs real-time image processing for embedded devices

Conventional CPU: Serial processing at pixel (dot) level / IMAPCAR: Parallel processing at line level

IMAPCAR Features

  • High performance: 170 GOPS(*1)
    Real-time image processing for robots and other embedded devices
  • All-software parallel processing using 32 to 128 processors
    Flexibly change algorithms during research and development
  • High reliability: Automotive quality design
    Support for everything from R&D to industrial applications with devices having track record in Automotive use

*1: Giga Operations per Second (GOPS) is the number of billions of operations performed in one second (performance for IMAPCAR2-200)


Overview of IMAPCAR Device (IMAPCAR2-200)

Parallel processing architecture enables fast image processing

  • 64 processing units (PEs) follow same instruction
  • Multiple instructions can be executed simultaneously in one cycle
  • Each PE has independent built-in local memory

IMAPCAR Block Diagram

Wide Application

IMAPCAR's practical image-recognition capability means that it has potential for a wide range of applications in vehicle safety systems, including lane-keep assist and pre-crash avoidance systems.



Inquiry Concerning Automotive LSI




End of this page.
Top of this page