A Closed-Loop Filtering Framework for Active Contours

本提案中描述的项目旨在开发一个用于生成鲁棒计算机视觉算法的框架，用于闭环系统。该研究的主要目标是研究控制理论在增强闭环计算机视觉算法中可能发挥的作用。作为一个传感器，成像系统可能会受到噪声的影响，无论是通过实际的传感过程还是通过成像场景的几何形状。因此，计算机视觉过程可以被解释为在存在噪声和不确定性的情况下的信号处理任务。这对于闭环视觉系统来说更加复杂，因为控制努力可以引入额外的干扰。通过对有限维系统的经典Luenberger观测器的分析，PI建议系统地建立一个类似的框架来过滤封闭曲线，这是被称为活动轮廓的计算机视觉算法的副产品。本质上，这项研究工作涉及到一个卡尔曼滤波算法的封闭曲线的发展，并涉及到一个原则性的预测和更新计划。主要的研究挑战在于这样一个事实，即封闭曲线形成一个无限维的空间，经典的状态空间观测器理论是无法处理的。将计算机视觉算法作为控制和动态系统的组成部分的成功考虑，能够极大地提高其鲁棒性水平，而不会严重复杂化其处理的性质。作为一个特殊的应用程序，展示了这一研究途径在特定领域的潜在影响和挑战，PI寻求研究在微观尺度上的生物膜的闭环控制和估计。本文描述的计算机视觉概念能够改善基于视觉的算法产生的信号，这些算法形成了闭环过程的主要信息路径。拟议的教育计划将其中许多想法纳入现有的计算机视觉课程，因为它们是将计算机视觉用于反馈驱动系统的重要组成部分。其次，预计夏季研究和高级设计项目将激发计算机视觉控制和动力系统理论的重要性。

The project described in this proposal seeks to a develop a framework for generating robustcomputer vision algorithms for use in closed-loop systems. A principal goal of the research endeavouris to examine the role that control theory may play in enhancing closed-loop computer visionalgorithms.Intellectual Merit. As a sensor, the imaging system can be wrought with noise, either throughthe actual sensing process or through the geometry of the imaged scene. Consequently, the computervision process can be interpreted as a signal processing task in the presence of noise and uncertainy.This is further compounded for closed-loop vision systems, because the control effort can induceadditional disturbances.Through an analysis of the classical Luenberger observer for finite- dimensional systems, the PIproposes to systematically build up a similar framework for filtering of closed curves, which are theby-products of th e computer vision algorithms known as active contours. Essentially, this researchendeavour involves the development of a Kalman filter algorithm for closed curves, and involves aprincipled predict and update scheme. The main investigative challenge lies in the fact that closedcurves form an infinite-dimensional space which classical state-space observer theory is not capableof handling.Broader Impact. The successful consideration of computer vision algorithms as components ofa control and dynamical system has the ability to tremendously increase their level of robustness,without severely complicating the nature of their processing. As a particular application demonstratingthe potential impact and challenges of this research avenue within a specific area, the PIseeks to study the closed-loop control and estimation of bio-membranes at the mico-scale. The observerconcept described herein has the capacity to improve the signal generated from vision-basedalgorithms that form the main information pathway of a closed-loop process.The proposed education plan incorporates many of these ideas into existing computer visioncourses, as they are essential components in the use of computer vision for feedback-driven systems.Secondly, summer research and senior-design projects are anticipated to motivate the importanceof control and dynamical systems theory for computer vision.