Improved Particle Filter Algorithm for Multi-Target Detection and Tracking

Area of Science:

• * Radar Systems Engineering
• * Signal Processing
• * Computational Intelligence

Background:

• * Particle filters are crucial for real-time target detection and tracking in radar systems, excelling in nonlinear and non-Gaussian environments.
• * Traditional particle filters struggle with complex dynamic scenes, leading to sample degradation, reduced accuracy, and difficulties in multi-target tracking.
• * Existing limitations hinder particle filter applications in complex, multi-target scenarios, necessitating advanced algorithmic solutions.

Purpose of the Study:

• * To develop an improved particle filter algorithm for robust multi-target detection and tracking in radar systems.
• * To enhance particle diversity and improve the efficiency and accuracy of particle replication for better state estimation.
• * To overcome the limitations of traditional particle filters in complex dynamic environments.

Main Methods:

• * Proposed an improved particle filter by dividing particles into 'tracking particles' for trajectory maintenance and 'searching particles' for target identification.
• * Integrated density-based spatial clustering with noise (DBSCAN) into the resampling phase to enhance particle replication efficiency and accuracy.
• * Developed a novel algorithmic framework to improve robustness and accuracy in complex multi-target scenarios.

Main Results:

• * The improved particle filter demonstrated enhanced particle diversity and more effective particle replication.
• * Achieved significant improvements in detection probability for multi-target scenarios.
• * Showcased a lower root mean square error (RMSE) compared to traditional methods.
• * Exhibited stronger adaptability and stable tracking capabilities in complex multi-target environments.

Conclusions:

• * The proposed improved particle filter algorithm effectively addresses the deficiencies of traditional methods in complex radar scenes.
• * The novel approach enhances the robustness and accuracy of multi-target detection and tracking.
• * This advancement holds significant potential for improving the performance of modern radar detection systems.