Teledyne FLIR Launches Prism ISP & Expands AI Capabilities

Prism ISP
Prism™ ISP, a highly optimized image signal processing software development kit (SDK) for embedded systems designed to maximize performance in thermal and multispectral applications.

Teledyne FLIR, a division of Teledyne Technologies, has unveiled the 1.0 version of its Prism™ ISP software development kit (SDK), known as Prism ISP. This software is designed for embedded systems and aims to maximize performance in thermal and multispectral applications, expanding its capabilities to low-power GPUs from Qualcomm and NVIDIA. Prism ISP offers various image-enhancing features such as noise reduction, super-resolution, electronic image stabilization, infrared (IR) and electro-optical (EO) video fusion, turbulence mitigation, and locally adaptive tone mapping.

The Prism ISP libraries are intended to empower integrators in the defense, commercial, and industrial sectors to significantly enhance image quality as part of artificial intelligence (AI) data capture at the edge. This not only improves performance but also reduces development costs and shortens the time to market. Developers, data scientists, and perception engineers can utilize Prism ISP with Teledyne FLIR’s leading thermal cores, including Boson®, Tau® 2, Hadron™, and Neutrino™.

In addition to Prism ISP, Teledyne FLIR has also enhanced Prism AI, its perception software. Prism AI enables classification, object detection, and object tracking in both thermal and visible light spectrums. It now offers high-performance models trained on large thermal image datasets for various applications, such as ground intelligence, surveillance, reconnaissance (ISR), air-to-ground, and counter-drone. This expansion complements the automotive software library introduced last year.

Dan Walker, Vice President of Product Management at Teledyne FLIR, noted that Prism ISP plays a critical role in the development of thermal-based AI systems. It offers exceptional performance while simultaneously reducing development costs. With Prism ISP and Prism AI, Teledyne FLIR aims to support a broader range of projects and custom developments that utilize low-power embedded processors at the edge, further extending their computational imaging ecosystem.

This development showcases Teledyne FLIR’s commitment to advancing signal processing and expanding AI capabilities at the edge, making it a valuable player in the thermal imaging and multispectral applications market.