It has been announced that Almalence has collaborated with Cadence to port their Video SuperSensor video image quality improvement software to the Cadence Tensilica Vision DSP.
When combined with the vision processing power of the Tensilica Vision DSP, the Almalence Video SuperSensor software can make image quality improvements at video framerates, previously only achievable on still images. Video SuperSensor can simultaneously improve resolution, low light sensitivity and dynamic range without any changes to the camera hardware and is useful for battery-powered embedded platforms such as mobile, drone, and Augmented and Virtual Reality (AR/VR) devices.
Tensilica Vision DSPs provide the necessary power and performance for running complex imaging algorithms efficiently, with lower power consumption and higher performance. After being ported to the Tensilica Vision DSP, Almalence Video SuperSensor was tested and shown to achieve up to 10x more performance in regards to frames per second with the Vision DSP over the use of a CPU and GPU alone, which lack the power and computational efficiency to run complex imaging software at 1080p frame rates inside practical battery and thermal constraints.
“We have experienced a very successful collaboration with the Cadence Tensilica team. The SIMD VLIW architecture of the Vision DSP enabled us to achieve a very efficient implementation of our Video SuperSensor software in a very short time,” said Eugene Panich, CEO at Almalence. “Due to the excellent software development tools and imaging library support provided by Cadence, it was remarkably fast and easy to port our software.”
“The form factor of mobile devices has made it very difficult until now to provide high-quality zoom in a low-light environment,” said Dino Bekis, Vice President, Product Marketing for the IP group at Cadence. “The combination of the Video SuperSensor software from Almalence and the Tensilica Vision DSP enables a highly efficient solution for battery-powered devices that significantly improves the user experience.”
The Tensilica Vision DSP family is designed for demanding imaging, computer vision and convolutional neural network applications in the mobile, automotive, surveillance, gaming, drone and wearable markets.