FA865124CB023

Small Technology Transfer Research Program (STTR) Phase II

Tactical Funding Increase for Improved Image Capture for AI Training Simulation

In particular, the project will include development of a custom stitch-free 360-degree camera system, capable of capturing 360 x 300 degree field of view spherical video, with 12 bit color, global shutter imaging, and high frame rate. The project will include the development of GPU algorithms to enable real-time image capture and livestreaming via a 5G network. The GPU developer must translate Circle Optics proprietary C code to the CUDA environment to accept a series of input pixels representing individual lens calibration data and generating a final lookup table by which pixels from every camera buffer are output to a specific pixel on the final output panorama. The Nvidia GPU processor will be responsible for moving pixels from each input buffer to a composite output buffer and correcting the pixel value to take into account color correction software. Exposure control will be implemented which reads in the average intensity of every camera buffer and modifies the exposure of the camera sensor until all cameras aggregate to a common average of pixel intensity values. This control will be handled over I2C communication between the CPU and each image sensor board. Furthermore, via a master clock, the CPU will provide a trigger signal to every camera at the same time so that image readout occurs on a common timing. The GPU will then aggregate and deliver the outgoing image data through a 5G wireless transceiver, beaming the data from the camera directly a URL that can be picked up by the dome theater.

The goal of phase II is to continue the R&D efforts initiated in Phase I. Funding is based on the results achieved in Phase I and the scientific and technical merit and commercial potential of the project proposed in Phase II. STTRs are completed in conjunction with a research institution.

AF21A-TCSO1

FX21A-TCSO1-0213

X21.A

Jared Evans