SBIR Phase II: Fast Panoramic Image Capture for Unmanned Aerial Vehicles

这个小型企业创新研究（SBIR）第二阶段项目的更广泛的影响/商业潜力是使无人机（UAV）在真实的房地产、农业、建筑、能源、娱乐、石油天然气和新闻等行业得到更广泛的采用。 该项目开发先进的无人机相机系统，具有无缝360度成像能力，以增加飞行时间和成像能力。最终用户受益于减少的劳动力需求和偏远地区的实时高保真图像，否则无法访问。工业检查、作物管理、田间测量和现场安全通常是通过将人员部署到田间来进行的，这通常会造成劳动力瓶颈和潜在的危险情况。UAV在这些应用中的采用可能受到图像处理的限制，这些图像处理对于UAV的使用来说消耗太多的功率，延迟的图像呈现影响导航能力，或者模糊与市场需求相关的细节的图像伪影。 这些限制都可以通过这里开发的相机技术来克服。这个小型企业创新研究（SBIR）第二阶段项目开发了一个轻量级的、无松弛的360度全景相机系统，该系统具有创新的光学设计，基于相邻相机的多面体形状配置。每个摄像头的视野紧密邻接但不与下一个重叠，从而能够从360度视图中无缝且计算简单地捕获图像，同时限制摄像头之间的盲区和丢失的内容。这种方法消除了与广角镜头和多相机系统相关的拼接误差和图像失真，并降低了360度图像捕获的计算需求。因此，低能量边缘计算以真实的时间呈现图像，而对无人机（UAV）应用中的飞行时间或能量使用的影响最小。此外，该项目还推进了针对无人机重量和尺寸进行优化的系统设计，并改进了运动空间框架的概念，其中相机可以集成在一起，形成强大的多相机系统。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact / commercial potential of this Small Business Innovation Research (SBIR) Phase II project is to enable wider adoption of unmanned aerial vehicles (UAVs) across industries such as real estate, agriculture, construction, energy, entertainment, oil & gas, and journalism. The project develops advanced UAV camera systems with seamless 360-degree imaging capabilities to increase flight time and the imaging capability. End users benefit from reduced labor demands and real-time high-fidelity images of remote areas otherwise inaccessible. Industrial inspections, crop management, field surveying, and site security are most often conducted by deploying humans into the field - often creating labor bottlenecks and potentially dangerous situations. Adoption of UAVs for these applications can be limited by image processing that consumes too much power for UAV use, delayed image presentation affecting navigation capabilities, or image artifacts that obscure the details pertinent to market needs. These limitations can all be overcome with the camera technologies developed here.This Small Business Innovation Research (SBIR) Phase II project develops a lightweight, parallax-free, 360-degree panoramic camera system with an innovative optical design based on a polyhedron-shaped configuration of adjacent cameras. Each camera's field of view closely abuts but does not overlap with the next, enabling seamless and computationally simplistic capture of images from 360-degree views, while limiting blind regions and lost content between cameras. This approach eliminates the stitching errors and image distortion associated with wide angle lenses and multi-camera systems, and reduces the computational demand for 360-degree image capture. As a result, low energy edge-computing presents images in real time with minimal impact on flight time or energy use in unmanned aerial vehicles (UAV) applications. Furthermore, this project advances system designs optimized for UAV weight and size, as well as improved concepts for kinematic space frames, in which the cameras can be integrated to form robust multi-camera systems.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.