Case Study: Designing Ultra HD Image Acquisition System, using Zynq UltraScale+ MPSoC Devices for Medical Imaging

System design and architecture

System designers required an image acquisition system that combines high-precision image sensors with high-speed digital processing capabilities. The image acquisition system with two MIPI cameras is designed to capture real-time images of the specimen (up to 8K UHD) and process them on the host PC. Advanced image analysis combined with machine learning & deep learning techniques is employed in the Host PC for accurate diagnosis.

The first step in the image acquisition process is to capture high-speed raw images from the sensors. The Xilinx Zynq UltraScale+ MPSoC platform is the heart of the system and is responsible for capturing high-resolution images from the MIPI cameras at 10Gbps link speeds. NVMe SSDs serve as a local frame buffer for data storage and retrieval. These SSDs deal with the difference in data rates between input and output. Finally, the captured data is delivered to the host PC via a 10G Ethernet link through the FTP protocol.

The embedded microcontroller on the board handles all functions, from capturing raw data from the cameras to transmitting it to the host PC for further analysis.

iWave took complete ownership of the design and development of the custom hardware, RTL development, and petalinux development reducing the design cycles. iWave also took on the responsibility of integrating third-party IP on the platform, to provide a complete data acquisition solution to the customer. A comprehensive ecosystem was provided around the hardware, software, and FPGA for the customer to quickly get started with the development of their final product.

Distinct advantages of using Zynq UltraScale+ MPSoC System on Module:

ZU7/5/4 SOM

• Multiple core ARM processors with flexible programmable logic cells
• Support for multiple connectivity standards – Ethernet, SATA, USB, HDMI, SDI, and more
• Reduce bottlenecks and latency through optimal hardware/software partitioning for image/video processing, analytics, and camera controls
• Integration of single-chip system processor and image processing reduces overall BOM cost and power consumption
• Optimize end-to-end latency through H.264/H.265 Video Codec Units

“Because medical devices should have a relatively long lifetime, with strong supply chain support and a minimum of 10+ years of product longevity provided by iWave, the product designers can focus on their design without worrying about the product going absolute.”

More information on the Zynq® Ultrascale+™ MPSoC development platform can be found here.

If you have any further queries, please contact us at mktg@iwavesystems.com.