Renesas and Hella Aglaia GmbH, a developer of intelligent visual sensor systems, have introduced an open, scalable front camera solution for driver assistance systems (ADAS) and automated driving.
The front camera solution combines Renesas' SoC R-Car V3M image recognition system with Hella Aglaia's camera software for Level 2 and 3 of the SAE automated driving scale. The solution enables system developers to devise a wide range of front-end cameras, up to level 3.
The cooperation between the chip manufacturer from Japan and the developer of intelligent visual sensor systems from Berlin gives the market for front view cameras a powerful boost, as the sensor vision is to become one of the most important requirements for NCAP in the next few years. According to Strategy Analytics, the front-end camera market is expected to exceed 200 million units over the next five years. This is made possible by the rapid introduction of NCAP safety functions such as automatic emergency braking, lane keeping warning and traffic sign recognition. “OEMs and Tier Ones are looking for open, but proven solutions with the right performance and cost-effectiveness to give them the freedom to implement a front camera with their own distinctive features," comments Strategy Analytics expert Ian Riches.
The Renesas R-Car V3M SoC implements an energy-efficient computer vision platform with various hardware accelerators, including a versatile Pipeline Engine (IMP) and a Computer Vision Engine (CVE), which allows the R-Car V3M to operate convolutional neural networks for the classification of objects and to manage algorithms such as optical flow and object recognition. The R-Car V3M also has an integrated ISP and a single DDR3L memory channel that offers further cost-cutting potential. The R-Car V3M is part of the Renesas Autonomy Platform, the chipmaker’s platform for ADAS and automated driving, providing end-to-end solutions from cloud to sensor and vehicle control.
EPTA Conference 2017 „Shaping the Future of Mobility“ Luzern, Verkehrshaus, Mittwoch, 8. ...»weiterlesen
Am 22. September war www.auto-mat.ch live vor Ort, als die ersten beiden automatischen ...»weiterlesen
Deutscher Verkehrsminister Dobrindt: Weltweit erste Leitlinien für Fahrcomputer»weiterlesen
Rapport du Conseil fédéral en réponse au postulat Leutenegger Oberholzer 14.4169 ...»weiterlesen