Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Visual Tracking with Intermittent Visibility: Switched Control Design and Implementation

Yangge Li, Benjamin C Yang, Sayan Mitra

Abstract

This paper addresses the problem of visual target tracking in scenarios where a pursuer may experience intermittent loss of visibility of the target. The design of a Switched Visual Tracker (SVT) is presented which aims to meet the competing requirements of maintaining both proximity and visibility. SVT alternates between a visual tracking mode for following the target, and a recovery mode for regaining visual contact when the target falls out of sight. We establish the stability of SVT by extending the average dwell time theorem from switched systems theory, which may be of independent interest. Our implementation of SVT on an Agilicious drone [1] illustrates its effectiveness on tracking various target trajectories: it reduces the average tracking error by up to 45% and significantly improves visibility duration compared to a baseline algorithm. The results show that our approach effectively handles intermittent vision loss, offering enhanced robustness and adaptability for real-world autonomous missions. Additionally, we demonstrate how the stability analysis provides valuable guidance for selecting parameters, such as tracking speed and recovery distance, to optimize the SVT's performance.

Visual Tracking with Intermittent Visibility: Switched Control Design and Implementation

Abstract

This paper addresses the problem of visual target tracking in scenarios where a pursuer may experience intermittent loss of visibility of the target. The design of a Switched Visual Tracker (SVT) is presented which aims to meet the competing requirements of maintaining both proximity and visibility. SVT alternates between a visual tracking mode for following the target, and a recovery mode for regaining visual contact when the target falls out of sight. We establish the stability of SVT by extending the average dwell time theorem from switched systems theory, which may be of independent interest. Our implementation of SVT on an Agilicious drone [1] illustrates its effectiveness on tracking various target trajectories: it reduces the average tracking error by up to 45% and significantly improves visibility duration compared to a baseline algorithm. The results show that our approach effectively handles intermittent vision loss, offering enhanced robustness and adaptability for real-world autonomous missions. Additionally, we demonstrate how the stability analysis provides valuable guidance for selecting parameters, such as tracking speed and recovery distance, to optimize the SVT's performance.

Paper Structure

This paper contains 16 sections, 1 theorem, 14 equations, 5 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Visual Tracking: Problem & Control Design
  3. Design of Switched Visual Tracking
  4. Computing Recovery Pose
  5. Stability of Target Tracking System
  6. Stability of Switching System
  7. Average Stable Dwell Time
  8. Stability Analysis
  9. Experimental Evaluation
  10. Experimental Setup
  11. Effectiveness of Tracking Algorithm
  12. Tracking Rate and Performance Tradeoff
  13. Recoverability-Performance Tradeoff
  14. Stable Dwell Time and Performance Tradeoff
  15. Related Works
  16. ...and 1 more sections

Key Result

Theorem 1

Suppose we have a collection of Lyapunov functions $V_p$$\forall p \in P_s$, and there exists $\mu>1, c>0$, such that for any even switch times $t_2,t_4,...$, Then, for any $\delta>0$, for any switching signal $\sigma$ with $\tau_{as}$ satisfy the the system is asymptotically stable with respect to

Figures (5)

  • Figure 1: The target (orange) moves from right to left and is initially visible to the pursuer (green). When the target exits the pursuer's view (red), the pursuer performs a backoff maneuver to regain visibility. The pursuer's field of view (FoV) angle, $\theta$, is represented by the triangle in front. The bottom row shows the camera images for three states of the pursuer.
  • Figure 2: Computation of the recovery pose $x_R$. The pursuer (black) at $x_R$ can include the whole reachable set of target (blue) in it's camera range.
  • Figure 3: The target drone (left) and the pursuer (right).
  • Figure 4: Sample of target trajectories Ellip (left) and SLem (right).
  • Figure 5: An example run of Ellip-1.5 using basline (left) and SVT (right). The pusuer is marked as red when the target is not visible.

Theorems & Definitions (4)

  • Definition 1
  • Definition 2
  • Theorem 1
  • proof