Automated laser tracking evaluation and scan constraint analysis for pose estimation (ALTESCAPE)

拟议的项目将开发使用主动激光雷达传感器对航天器和类似物体进行智能实时跟踪的新方法。近几十年来，主动传感器已经成为传统CCD相机的替代品，避免了对广泛地面支持和飞行前规划的需要。我们与瑞尔森大学、加拿大航天局和Neptec设计集团的研究人员合作，将解决目前阻碍激光雷达系统的几个操作和算法挑战。根据该计划开发的新技术将在Neptec的TriDAR跟踪系统上进行测试。这些创新将改善系统性能，提高可靠性并扩大TriDAR工业应用的范围。不受环境光照条件变化的影响，自发光有源传感器用激光束扫描物体，并提供3D距离数据用于姿态估计。主动传感器对于航天器之间的近距离操作特别有用。这种类型的传感器可以测量相对位置、方向和速度。该项目的主要目标是：目标1 -应用代数约束分析(a)帮助任务规划者选择有效的扫描区域和模式，以及(b)提供跟踪准确性的一致实时评估；目标2 -基于卡尔曼滤波（KF）和其他信号处理技术的运动适应姿态估计的发展，以在(a)可能涉及快速运动以及(b)目标1中提到的静态跟踪精度有限的情况下传播姿态。拟议的项目将有助于HQP的培训，并将增加加拿大大专院校对加拿大空间计划感兴趣的先进科学和技术研究领域的参与。开发的基础设施将吸引来自大多伦多地区其他大学的研究，从而有可能获得真实的距离数据，以验证计算机视觉算法的性能。

The proposed project will develop novel approaches to intelligent real-time tracking of spacecraft and similarobjects using active LIDAR sensors. In recent decades, active sensors have become an alternative toconventional CCD cameras, avoiding the need for extensive ground support and pre-flight planning. Ourcollaboration between researchers at Ryerson University, the Canadian Space Agency, and Neptec DesignGroup will address several operational and algorithmic challenges that currently hamper LIDAR systems. Newtechnologies developed under this program will be tested on Neptec's TriDAR tracking system. Theseinnovations will improve system performance, increase reliability and expand the range of TriDAR industrialapplications. Immune to variable ambient lighting conditions, self-illuminating active sensors scan an objectwith a laser beam and deliver 3D range data for pose estimation. Active sensors are particularly useful forclose-range operations between spacecraft. Sensors of this type can measure relative position, orientation andspeed.The key goals of the project are: Goal 1 - the application of algebraic constraint analysis to (a) aid missionplanners in the selection of productive scanning area and patterns, and (b) provide a consistent real-timeassessment of tracking accuracy; and Goal 2 - the development of motion-accommodating pose estimationbased on Kalman Filtering (KF) and other signal processing techniques to propagate pose when (a) fast motionmay be involved, and (b) static tracking accuracy, as referred to in Goal 1, is limited. The proposed project willcontribute to HQP training and will increase the involvement of Canadian post-secondary institutions inadvanced science and technology research areas of interest to the Canadian Space Program. The developedinfrastructure will attract researches from other universities across the Great Toronto Area giving possibility toobtain real range data to validate the performance computer vision algorithms.