Imaging and vision platform chosen for ADAS vision system

Kyoungsoo Kim, CEO of Nextchip commented: “We developed the APACHE4 to provide affordable and scalable ADAS systems for the mass market. CEVA’s industry-leading vision platform adds a high degree of flexibility to our solution to enable differentiated, machine vision-related ADAS products. CEVA’s long-standing experience and success in computer vision and artificial intelligence makes them the ideal partner for Nextchip to address new generations of ADAS systems.”

APACHE4 is a vision-based pre-processor SoC targeting next-gen ADAS systems. With a dedicated sub-system of image processing accelerators and optimised software, APACHE4 dramatically reduces the main ECU’s workloads by as much as 70% with improved performance, to allow all detection algorithms to operate simultaneously. The APACHE4 incorporates dedicated detection engines that include pedestrian detection, vehicle detection, lane detection and moving object detection. The embedded CEVA-XM4 imaging and vision platform enables customers to develop differentiated ADAS applications by means of high-level software programming.

Gideon Wertheizer, CEO of CEVA commented: “Nextchip’s selection of the CEVA-XM4 is another endorsement of the unique capabilities of our vision IP and its value proposition for the ADAS market. The APACHE4 brings key ADAS features within reach of every car manufacturer and model, moving us closer to a world of enhanced driver safety and, ultimately, safer roads.”

CEVA’s latest generation imaging and vision DSP platforms address the extreme processing requirements and low power constraints of the most sophisticated machine learning and machine vision applications used in smartphones, surveillance, augmented reality, sense, avoid drones and self-driving cars. These DSP-based platforms include a hybrid architecture composed of scalar and vector DSP processors coupled with a comprehensive Application Development Kit (ADK) to streamline software deployment. The CEVA ADK includes: CEVA-Link for seamless software level integration with a host processor; a range of widely used and optimised software algorithms; the CEVA Deep Neural Network (CDNN2) real time Neural Network software framework which streamlines machine learning deployment at a fraction of the power consumption of the leading GPU-based systems; and development and debugging tools.

CEVA and Nextchip will attend the AutoSens Conference, as part of AutoWorld, in Brussels, 19th to 21st September 2017.