The industry-leading ultra-low power CNN IP supports the AI processing functionality required for automotive front view cameras.
On March 2, 2018, Japan Tokyo, Japan - the world's leading automotive semiconductor solution Supplier Ruisa Electronics Co., Ltd. today announced that the new R-Car V3H chip system (SOC) is launched. The SoC leads to the industry's leading low-power, providing powerful computer visual performance and artificial intelligence processing capabilities for automotive front-view cameras, suitable for mass production 3 (conditional automation) and level 4 (highly automated) automatic driving car note 1. The new R-Car V3H SOC is optimized for three-dimensional front-view camera applications, and its computer visual performance announced on April 11, 2017, the R-Car V3M SOC of NCAP Note 2 front-view cameras. As an open, innovative, reliable renesas AutonomyTM Platform, an automatic driving assistance system (ADAS) and automatic driving (AD), R-Car V3H Helps Level 1 supplier and OEM flexibly achieve their auxiliary driving to high automatic driving The road to the car.
Advanced R-CAR identification technology
R-Car V3H SOC focuses on architectural optimization of computer vision processing, supports all ADAS related functions from conditional automatic driving to high automatic driving. R-Car V3H uses Rissa based on the IMP-X5 + image recognition engine and the heterogeneous computer visual kernel concept of the dedicated hardware accelerator, and is implemented with algorithms including fused light notes 3, dense stereoscopic 4 and object classification Note 5. Advanced perceived function. Integrated CNN Note 6ip speeds up deep learning with only 0.3W industry, with a deep neural network performance of 2 times R-Car V3M.
Scalability and cost savings
The R-Car V3H is based on the verified, which is used for IP of R-Car V3M, and converts the camera sensor signal with a dual channel image signal processor (ISP) to establish an image and perform identification processing. This IP reuse ensures that the development time is saved from the 3-level and 4-level smart camera system using R-Car V3M, saves development time, and saves the development time, while making the system no longer requires each camera. ISP has reduced costs. In order to further save system cost, R-Car V3H uses only a single LPDDR4 memory, so that the cost of the memory element is reduced compared to other front-view camera solutions.
Open platform solution
Level 1 Suppliers and OEMs can choose to develop your front-view camera solutions, or work with leading partners from the RenesAs Autonomy ecosystem, the ecosystem has been supported by many of Hella Aglaia.
Jean-Francois Chouteau, Vice President, Renesa, said: "The specifications and design of R-Car V3H are completed by close cooperation with the front-view camera market leaders to ensure the needs of the leading automatic driving system innovation technology. As automotive semiconductor Leaders in the industry, Rissisa electronics is committed to providing open, innovative and reliable solutions for assist driving and automatic driving. In addition to achieving higher computer visual performance with very competitive system costs, our customers use R-Car In V3H, especially like it is possible to protect the freedom of front-view cameras with its own unique differentiation technology, while still benefiting from R-Car V3m solutions to R-Car V3H solutions. "
Supply
Samples of R-Car V3H SOC will start from the fourth quarter. The mass production plan began in the third quarter of 2019. (The delivery date is subject to change without notice.)
Renesa showed new R-Car V3H on Embedded World 2018 held in Nuremberg, February 27, 2018, showing new R-Car V3H, booths in the Exhibition Hall 310.
If you need to understand the product specifications of the R-Car V3H, please check the information attached to the text.
Note 1 SAE International's 3-level (conditional automation) and level 4 (highly automated) in new standard J3016
The classification and definition of the type of automatic driving system on the road provides a unified classification system and support definition, from "non-automated" to "fully automated", specifies 6 driving automation levels.
Note 2 NCAP (new car assessment plan)
A government-developed automotive safety plan used to assess the ability of new automotive design to resist various safety threats.
Note 3 dense light flows to accurately track objects flow.
Note 4 dense three-dimensional parallel use with all the pixel units of the camera image, measure distance in a 3D manner.
Note 5 The object classification is used to identify objects.
Note 6 Convolutional Neural Network (CNN) is a depth, feedforward artificial neural network, has been successfully used to analyze visual images, and increasingly increasingly used in automotive applications such as road detection, subject classification.
About Reisa Electronic Co., Ltd.
Renesa Electronic Co., Ltd. (TSE: 6723) provides professional trusted innovative embedded design and complete semiconductor solutions to improve people's work and life by using billions of net intelligent equipment to use their products. Way. As a leader in the world's leading microcontrollers, analog power devices and SOC products, Renesa Electronics is available for automobiles, industrial, home (HE), office automation (OA), information communication technology (ICT) and other applications Professional technical support, quality assurance and comprehensive solutions, look forward to working with you to create unlimited future. Read more
Our other product:
