• A platform that supports image pipeline processing performance, image enhancement, and embedded vision functions, including shake compensation, face recognition, gesture recognition, and augmented reality.
• CEVA-MM3101 gesture recognition function demonstrated at CES2012

January 11, 2012 - CEVA (NASDAQ: CEVA, LSX: CVA), a leading licensor of Silicon Intellectual Property (SIP) platform solutions and DSP cores, today announced the first imaging and vision platform based on the CEVA-MM3000™ architectural framework. The CEVA-MM3101™ is a fully programmable, low-power platform designed to address the advanced image processing needs of camera-embedded devices such as smartphones, tablets and smart TVs.

The CEVA-MM3101 integrates video and imaging functions previously distributed across multiple processing engines, handling video and imaging pipelines, image enhancement, embedded vision applications, and image encoding. Rather than relying on fixed-function engines or leveraging the main application processor, the CEVA-MM3101 supports all these functions on a single processor, resulting in a 20x power savings compared to CPU-based solutions.

The particularly powerful CEVA-MM3101 enables new levels of high-performance image enhancement and embedded vision applications previously impossible with RISC CPU-based designs. For example, the CEVA-MM3101 can process 1080p video streams or 8 megapixel images at 12 frames per second while simultaneously performing advanced features such as image stabilization, color correction, wide dynamic range (WDR), and face/gesture recognition.

“With the emergence of new video, imaging and embedded vision applications, high-performance DSP technology is playing a critical role in mobile and digital home applications,” said Gideon Wertheizer, CEO of CEVA. “Traditional hardwired blocks and general-purpose CPUs no longer provide the flexibility, power efficiency or performance required to keep pace with rapidly evolving market trends. The CEVA-MM3101 targets these complex applications and, through software integration, provides our customers with a cost-effective, low-power and differentiated solution.”

Embedded Vision Sets New Milestones
CEVA is collaborating with eyeSight, a CEVAnet partner, to enter the emerging embedded vision space. This collaboration will deliver a comprehensive Human Machine Interface (HMI) based on the CEVA-MM3101 platform. EyeSight tracks multiple users' fingers in a variety of environments, enabling gesture recognition and a virtual mouse user interface. CEVA and eyeSight will demonstrate a variety of real-world applications utilizing this technology at CES 2012.

“We congratulate CEVA on developing a powerful platform that integrates imaging and vision technologies,” said Jeff Bier, founder of the Embedded Vision Alliance ( http://www.Embedded-Vision.com ). “Next-generation mobile and consumer devices are rapidly adopting embedded vision capabilities, including gesture-based user interfaces and object recognition for applications such as augmented reality gaming, security and navigation,” said Dr. Roberto Año, CEVA’s vice president of product development. “Implementing this type of visual intelligence in consumer products requires a combination of high processing performance, programmability, energy efficiency and low cost. The introduction of a licensable processing subsystem specifically targeted at embedded vision applications is a significant step forward for CEVA and the industry.”

Development Tools and Support
The CEVA-MM3101 platform provides a software development toolkit, including an optimized C compiler, IDE, debugger, simulator, and profiler. The user-friendly IDE facilitates developer implementation, debugging, optimization, and code execution. The IDE provides information about all cores and memory resources to support vector processor optimization. The solution also includes a comprehensive library of video and image functions, including image pipeline kernels, linear and nonlinear filters, pre- and post-processing functions, face recognition, and video codec kernels.
More information about the CEVA-MM3101 can be found at www.ceva-dsp.com/ISP .