Smart Embedded Vision is opening up new possibilities for implementing stand alone or edge of net systems to analyse image data and make decisions across a broad spectrum of uses. Applications such as, drones, machine vision, thermal imaging, gaming, video surveillance, robotics, Advanced Driver Assistance Systems (ADAS), machine learning and HMI all require processing, often in real time, of image data camera inputs and displays. Critical in these applications is the security and integrity of the data being processed and secure high-speed interfaces for communication and transmission of the data.
The inherent parallel processing and high-speed I/O capabilities of FPGAs makes them ideal processing platforms for delivering the high data throughputs needed for both high-resolution imaging and machine learning algorithms. Microchip provides FPGA imaging and video solutions to enable the evaluation of multiple protocols and the development of a host of image and video processing applications. Microchip PolarFire™ FPGAs are ideal for HD and UHD (4K/2K) imaging/video applications due to their rich memory and on chip digital signal processor (DSP) resources.
Microchip’s solutions come with a complete ecosystem, including comprehensive application-specific hardware, optimized intellectual property suite for image processing, sample reference designs, demonstration designs and collateral. The comprehensive IP portfolio covers the latest embedded vision Ips such as MIPI, SLVS-EC, CoaXPress, Serial Digital Interface (SDI), HDMI, etc, while our partners cover offerings in Machine Learning, H.264, DisplayPort and HDCP IPs, with many more IPs being added continuously. Using the Microchip comprehensive development tools it is relatively easy to build the IP blocks into a complete imaging solution
Many embedded vision systems require a more detailed analysis of the image data local to the data source. Applications such as smart machine vision and video security need some intelligent image analysis. Microchip offer the VectorBlox Accelerator Software Development kit for the Polarfire FPGA and Polarfire SoC FPGA devices. The VectorBlox Accelerator SDK contains different tools that compile a neural network description from TensorFlow and ONNX into a Binary Large Object (BLOB) that is stored in flash and loaded into the DDR memory during execution. This can be used to build intelligent vision systems with suitable training material. To aid development Microchip have a Vectorblox Demonstration available running on the PolarFire Video evaluation kit.
For more information on Microchip FPGAs and Smart embedded imaging solutions please refer to our product webpages or