Detroit, Michigan, October 23, 2014 – Videantis and VISCODA, both leaders in video analytics solutions for automotive ADAS systems, announced today they have joined forces and are demonstrating key automotive vision algorithms and processing technology at the Ethernet & IP @ Automotive Technology Day in Detroit on October 23 and 24.
VISCODA’s history is in algorithm development for digital image processing, computer vision, and 3D scene analysis, while videantis licenses scalable processors optimized for computer vision and video coding, bundled with optimized software libraries.
One unique feature of the combined solution on display is that the demanding ADAS algorithms don’t run on a high-end multicore CPU/GPU combination in the dashboard. Instead, the algorithms run on a very small, low-power videantis processor integrated inside the camera. Such a setup has two distinct benefits. First, with each camera that is added to a car, the available compute power increases, so the number of cameras will never create a processing bottleneck. Second, the system seamlessly integrates with Ethernet AVB, which significantly reduces cabling cost in the car.
Ethernet AVB does require an additional video compression step to transmit the video from the camera to the head unit. This hinders ADAS algorithms that run after the lossy compression step, which is the standard method. The newly announced solution runs the ADAS algorithms on the original images inside the camera, before they’re compressed. As a result, the ADAS algorithms don’t suffer from the artifacts introduced during compression and instead retain their accuracy. The videantis unified video/vision solution runs both the video analysis and the H.264 low delay video compression on the same processor architecture.
“We’re excited to work with videantis on bringing our complex vision algorithms to market. Their scalable v-MP4000HDX processor architecture provides 100x more performance than CPUs or GPUs and reduces power consumption by 1000x,” says Dr. Hellward Broszio, VISCODA’s CEO. “It was easy to get our algorithms up and running on the videantis platform, resulting in a solution that can quickly be integrated into the customer’s silicon.”
Dr. Hans-Joachim Stolberg, videantis’s CEO, says, “We launched our partner program last year, and we’re now very pleased to add VISCODA. They’re well known in the market for their key enabling, high-quality algorithms for automotive ADAS and consumer applications. Automotive OEMs, Tier 1s, and semiconductor suppliers can now start validating our joint solution, which is already running at speed on our silicon-based demonstrator.”
Videantis has the following technologies on display at the show:
- Structure from motion for depth extraction using a single camera
- Low-delay, 10/12-bit video codecs for Ethernet AVB
- Pedestrian detection using histogram of oriented gradients descriptors
- High-performance, scalable, licensable video/vision processor IP
The Ethernet & IP @ Automotive Technology Day is taking place at the Cobo Center in Detroit, Michigan from October 23 to 24. Videantis is located in booth 304. To request a meeting, please email sales@videantis.com or call us at +49 (511) 762-5046.
About videantis
Headquartered in Hannover, Germany, videantis GmbH is a one-stop video and vision processor IP and solution provider delivering flexible multi-standard HW/SW video, imaging, and vision solutions for mobile, consumer, and automotive markets. Based on a unified processor platform approach that is licensed to chip manufacturers, videantis provides tailored solutions to meet the specific needs of its customers. With their core competencies of deep video application expert know-how and strong SOC design and system architecture expertise, videantis passionately serves a worldwide customer base with a diverse range of target applications, such as mobile phones, advanced driver assistance systems, in-car infotainment, and over-the-top TV.
For more information about videantis, please visit www.videantis.com.
