We offer to accelerator solutions to achieve next-generation high-performance computer vision for such applications as quality assurance, manufacturing, smart mobility, medical imaging, robotics, and others.
The Embedded Systems Development Team from the Institute of Electronics and Computer Science of Latvia (EU) offers accelerator solutions to achieve next-generation high-performance computer vision for applications such as quality assurance, manufacturing, UAVs, medical imaging, robotics, and others.
We achieve higher performance than conventional CPU/GPU solutions can offer by utilizing Multi-Processor System-on-a-Chip (MPSoC) devices with on-chip Field-Programmable Gate Arrays (FPGA). FPGAs with their inherent parallel architecture and built-in digital circuit elements can be tailored to perform specific tasks, therefore providing a use-case specific digital circuit for the job
Lorem ipsum dolor sit amet, consectetur adipiscing elit.
During our research in computer vision, we have developed a collection of specialized components, e.g. lens distortion correction, perceptive transformation, feature extraction. These components are made as building blocks with a common data exchange interface, allowing for easy interconnection between them.
We pick from our ready-made building blocks and (if needed) design the missing ones to create an image processing pipeline for your specific task/use-case, therefore reducing the design time and costs.
We design an accelerator platform tailored for your current computer vision system with a suitable communication interface (PCIe, USB, Ethernet) for easy integration or we can design a new, standalone solution for your needs.
We deliver you the designed accelerator platform with the necessary software libraries and the documentation, which ensures straightforward and simple integration in your existing image processing system.
Lorem ipsum dolor sit amet, consectetur adipiscing elit.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Soluta postea epicuri cum ne, vide intellegebat est no, id duis suscipiantur eos.
Tailored processing pipeline for your application
Large image (20+ Mpix) processing in real-time (<20 ms per processing pipeline)
Image transformation and processing with performance close to a dedicated chip
Low power consumption (3+ times more efficient when compared to GPU implementations) with increased performance (suitable for battery-powered devices)
Dzerbenes str. 14, Riga, LV-1006, Latvia
silhouse@edi.lv
+ 371 67554500
