SBIR Phase II: Three Dimensional Monocular Thermal Ranging Camera

点击翻译按钮获取中文摘要

This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2). The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase II project is to improve the safe operation of Advanced Driver-Assistance Systems (ADAS) equipped vehicles and robotic mobility machinery while helping overcome the sensing challenges that have delayed autonomous vehicle (AV) adoption and acceptance. Degraded Visual Environments (DVE) such as solar glare, darkness, smoke, and dust impact the safety of vehicle occupants, as well as those external to the vehicle, especially pedestrians and cyclists. The new sensor modality of the 3D monocular thermal technology developed as part of this project may improve the safe operation of on-road and off-road vehicles alike. The system seeks to detect and locate pedestrians, cyclists and objects in all conditions, improving vehicle automated braking, obstacle avoidance, localization, and navigation. Beyond improving safety, the new 3D monocular thermal modality devise will improve the utility and commercial value of vehicles and robotic machines by extending the operating range to include the most challenging environmental conditions.This Small Business Innovation Research (SBIR) Phase II project will develop a new sensor modality that addresses the shortcomings of camera, radar and LiDAR sensors used in current automotive and robotic mobility applications. This project seeks to develop the simple, yet powerful innovation of thermal monocular ranging where a single thermal camera produces 2D video detail concurrent with a range map or point cloud of much higher spatio-temporal density than contemporary sensors. Drawing inspiration from monocular convolutional neural network (CNN) ranging work recently published in the visible domain, this project creates a new field of monocular 3D thermal imaging. Unlike visible-based CNN systems that rely on color information and reflected light, this project will construct from first principles new thermal networks and ground truth sets to train the CNN. The extreme price pressures of the automotive market can only be met through adoption of a new thermal detector technology. This project further develops new thin-film detector technology and the detector to readout interconnect application process with potential breakthroughs in thermal camera performance, utility, and affordability.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.