Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Dynamic Network Flow Optimization for Task Scheduling in PTZ Camera Surveillance Systems

Mohammad Merati, David Castañón

TL;DR

The paper addresses efficient scheduling of PTZ cameras in dynamic, crowded surveillance environments. It merges Kalman-filter-based motion prediction with a time-expanded dynamic network flow formulation to assign cameras to track and fixed locations, maximizing observation value. A greedy set-cover-based group-tracking extension reduces redundancy and computational load while preserving coverage. Experimental simulations against a master-slave baseline show improved coverage and reduced waiting times, with demonstrated scalability for real-time planning in dynamic scenes.

Abstract

This paper presents a novel approach for optimizing the scheduling and control of Pan-Tilt-Zoom (PTZ) cameras in dynamic surveillance environments. The proposed method integrates Kalman filters for motion prediction with a dynamic network flow model to enhance real-time video capture efficiency. By assigning Kalman filters to tracked objects, the system predicts future locations, enabling precise scheduling of camera tasks. This prediction-driven approach is formulated as a network flow optimization, ensuring scalability and adaptability to various surveillance scenarios. To further reduce redundant monitoring, we also incorporate group-tracking nodes, allowing multiple objects to be captured within a single camera focus when appropriate. In addition, a value-based system is introduced to prioritize camera actions, focusing on the timely capture of critical events. By adjusting the decay rates of these values over time, the system ensures prompt responses to tasks with imminent deadlines. Extensive simulations demonstrate that this approach improves coverage, reduces average wait times, and minimizes missed events compared to traditional master-slave camera systems. Overall, our method significantly enhances the efficiency, scalability, and effectiveness of surveillance systems, particularly in dynamic and crowded environments.

Dynamic Network Flow Optimization for Task Scheduling in PTZ Camera Surveillance Systems

TL;DR

The paper addresses efficient scheduling of PTZ cameras in dynamic, crowded surveillance environments. It merges Kalman-filter-based motion prediction with a time-expanded dynamic network flow formulation to assign cameras to track and fixed locations, maximizing observation value. A greedy set-cover-based group-tracking extension reduces redundancy and computational load while preserving coverage. Experimental simulations against a master-slave baseline show improved coverage and reduced waiting times, with demonstrated scalability for real-time planning in dynamic scenes.

Abstract

This paper presents a novel approach for optimizing the scheduling and control of Pan-Tilt-Zoom (PTZ) cameras in dynamic surveillance environments. The proposed method integrates Kalman filters for motion prediction with a dynamic network flow model to enhance real-time video capture efficiency. By assigning Kalman filters to tracked objects, the system predicts future locations, enabling precise scheduling of camera tasks. This prediction-driven approach is formulated as a network flow optimization, ensuring scalability and adaptability to various surveillance scenarios. To further reduce redundant monitoring, we also incorporate group-tracking nodes, allowing multiple objects to be captured within a single camera focus when appropriate. In addition, a value-based system is introduced to prioritize camera actions, focusing on the timely capture of critical events. By adjusting the decay rates of these values over time, the system ensures prompt responses to tasks with imminent deadlines. Extensive simulations demonstrate that this approach improves coverage, reduces average wait times, and minimizes missed events compared to traditional master-slave camera systems. Overall, our method significantly enhances the efficiency, scalability, and effectiveness of surveillance systems, particularly in dynamic and crowded environments.
Paper Structure (13 sections, 18 equations, 3 figures, 2 tables)

This paper contains 13 sections, 18 equations, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Problem Formulation
  3. Main Approach
  4. Group-tracking Task Assignment
  5. Value System
  6. Experimental Study
  7. Simulation Scenarios
  8. Evaluation Metrics
  9. Results
  10. Scenario 1
  11. Scenario 2
  12. Discussion
  13. Conclusion

Figures (3)

  • Figure 1: H-period online video surveillance network model with $T=5$. The graph includes $l$ camera nodes, $n$ track location nodes, and $m$ fixed location nodes, each duplicated across H periods. Node $D_i$ denotes the demand node for fixed locations which are repeated $H/T$ times. For simplicity, the $t$ is omitted in designating node $(X_i,t)$.
  • Figure 2: Performance of different systems in a scenario with 400 pedestrians and pedestrian generation with Poisson rate of 1/20 per frame. (a) and (c) show the ratio of covered and missed objects which must add up to one. We can see that the flexible system with group tracking has a ratio of covered objects close to 1 (only one lost object for this system while the flexible system lost 3), and this metric is 0.75 for the master-slave system. (b) shows the average wait time for the systems.
  • Figure 3: Performance of different systems in a scenario with 450 pedestrians and pedestrian generation with Poisson rate of 1/18 per frame. Only one object is lost for the flexible system with group tracking while the performance of the other systems dropped dramatically.