Xfuse, LLC 推出Phoenix ISP 并可立即下载AMD Kria KV260 Vision AI 入门套件

Artificial intelligence and computer vision engineers seeking real-time, data-rich HDR video to fuel their applications can now download the Phoenix ISP for the Kria KV260 Vision AI Starter Kit from the AMD Xilinx App Store

Burlingame, C.A. – February 2, 2023 – Today, Xfuse, LLC, a developer of customizable imaging and video technology, announced the immediate availability of its Phoenix HDR ISP for the AMD Kria™ K26 System-on-Module (SOM) and KV260 Vision AI Starter Kit. For AI developers and computer vision engineers needing to capture real-time, data-rich imaging streams, the Phoenix HDR ISP can deliver fully-tone mapped HDR video up to 120dB dynamic range even in high contrast lighting conditions. While the standard Phoenix configuration is now easily accessible through the AMD Xilinx App Store for Kria applications, Xfuse is also able to fully customize its imaging pipeline for the specialized needs of developers for other AMD adaptive devices, including Zynq™ 7000 SoC and Zynq UltraScale+™ MPSoC.

For step-by-step instructions on how to incorporate the Phoenix HDR ISP into their workflow, developers can visit the Xfuse website, here https://xfuse.ai/phoenix-isp-for-kria-som-2/.

In conjunction with the Phoenix HDR ISP’s immediate availability on the AMD Xilinx App Store for Kria applications, Xfuse is also now shipping its new Imager Access System (IAS) HDR sensor module tuned for the AMD Kria KV260 Vision AI Starter Kit. Based on the ON Semiconductor AR0239 CMOS imager capable of capturing up to 120dB/20EV of dynamic range at 1080p/30 full HD resolution, the sensor data is merged and tone mapped in the ISP leveraging Xfuse’s proprietary human vision technology. The Xfuse imaging module is ready for quick deployment in challenging environments encountered in AI, security, robotics, or consumer applications. The sensor module features an ON Semi AR0239 CMOS imager, an M12 (S-Mount) lens holder, and a 2.9mm manual focus lens (included) with a 128.7° field of view.

“Our Kria SOMs and Vision AI Starter Kit are designed to streamline the development of advanced vision applications,” said Chetan Khona, Sr. Director, Industrial, Vision, Healthcare & Sciences at AMD. “With the unique capabilities of Xfuse’s Phoenix Soft HDR ISP available on the Kria KV260 Starter Kit, our customers creating the next generation of smart city, security and retail vision systems gain access to highly-accurate, real-time data streams in our out-of-the-box development kit.”

Fully-Customizable Sensor Agnostic HDR Video

Adapted from Xfuse’s patented HDR technology, the Phoenix ISP is modeled on human vision. The proprietary technology guarantees retention of local image contrast as well as highlight and shadow detail, without producing halos or color shifts. Phoenix is also uniquely adaptable to support sensors of any resolution, color filter arrays and diverse HDR capture methods.

Xfuse is able to customize its full, end-to-end ISP to meet the unique needs of any specific application. For specialized mission critical applications requiring real time imaging data, Xfuse can fine-tune its proprietary advanced algorithms to accurately tone map high contrast scenes with very low latency.

“We are excited that our Phoenix Soft ISP solution is now available for immediate download to the developers operating within the AMD ecosystem,” said Alfred Zee, CEO of Xfuse. “The Kria SOM platform is helping make these advanced technologies available to a wider community of engineers. The benefit of a “Soft ISP” over a hardwired ISP means we are able to fully-customize our ISP to meet any specific application requirements.”

Technical Capabilities of the Phoenix HDR ISP

For engineers and camera designers developing mission critical applications, the highest-quality video signal and ultra-fast response times are imperative. Phoenix delivers real-time video capable of producing up to 120dB or 20EV steps of dynamic range while ensuring extremely low latency, and its unique HDR IP completely eliminates halo artifacts and color shifts. This allows Phoenix to capture up to 1080p HDR video in real-time that is fully tone-mapped at 30 fps. Phoenix is also able to accommodate automatic or full-user control modes.

Phoenix ISP Key Features

• Optimized for AMD Kria K26 SOM and KV260 Vision AI Starter Kit– Support for ON Semi AR0239 HDR 1080p/30 Full HD Sensor
• 3-frame exposure merge block
• Up to 120dB/20EV High Dynamic Range
• Proprietary locally adaptive tone mapping technology
• Auto/manual exposure & white balance
• Ultra Low latency
• Auto de-ghosting and motion compensation
• Auto halo elimination
• Auto EV bracketing

Pricing and Availability

Developers interested in adding the Phoenix HDR ISP to their development tool kit, for the AMD Kria K26 SOM, KV260 Vision AI Starter Kit, or in a customized configuration, can visit the Xfuse website for a step-by-step guide to downloading a free fully-functional watermarked trial version. For more information on the Phoenix ISP, to access reference designs, or download the ISP, please visit https://xfuse.ai/. Engineers and designers looking to explore integration partnerships or schedule a demo, please email info@xfuse.ai.

Developers interested in purchasing the plug-and-play Xfuse Image Sensor Module (AR0239) for the Phoenix HDR Image Signal Processor can purchase the device for $129 on Amazon.

About Xfuse, LLC

Headquartered in Burlingame, C.A., Xfuse, LLC is a developer of customizable imaging and video technology powering the next generation of artificial intelligence applications. The company’s proprietary High Dynamic Range (HDR) Image Signal Processing (ISP) technology provides developers with complete control over the critical imaging pipeline necessary to advance self-aware robotics, autonomous guided vehicles, machine vision, and more. Able to fuse numerous data streams from imaging and non-imaging sensors in real-time, Xfuse ISPs are available to both in-house development teams and independent developers alike. For more information on Xfuse, LLC please visit: https://www.xfuse.ai/.