Developing Live Real-Time Video Applications for Snapshot HSI

Snapshot hyperspectral imaging, which captures a low-resolution hyperspectral cube with every frame, can be integrated with a 5Mpixel RGB video stream for real-time scene analysis, combining spatial and spectral data. Hyperspectral cameras can be hardwired to a GPU to enable low-latency edge computing, wherein devices in remote locations can process data at the edge of the network. By capturing both modalities, synchronously and coaxially, traditional computer vision techniques (such as object detection, tracking and reacquisition) can be augmented with spectral analysis, improving accuracy in applications.

In this presentation, real-time applications across three domains will be demonstrated: patient triage for healthcare by rapidly examining physiological parameters; camouflage detection and situational awareness for security; and edge-based processing in the field to enhance decision-making in plant health monitoring (such as water scarcity or nutrient deficiencies) and crop yield estimation.

About the presenter
Nikhil Jawade is a field applications engineer at Living Optics, where he drives the exploration and development of practical applications for snapshot hyperspectral imaging. He specializes in designing and executing lab-based proof-of-concept experiments and translating them into robust, real-time applications. With over four years of experience in software engineering, and academic credentials including a degree in Computer Science and a Master’s in Artificial Intelligence, Jawade brings a unique blend of technical depth and applied innovation to the field of computational imaging.