Onboard Perception and Localization for Resource-Constrained Dynamic Environments: A RoboCup Small Size League Case Study

Self-localization is a core capability in autonomous mobile robot navigation. The Monte Carlo Localization (MCL) algorithm addresses this challenge by maintaining a particles set to represent multiple hypotheses of the robot’s current state. In 2022, this problem was introduced in the RoboCup Small Size League (SSL), and this work proposes an integrated pipeline for solving the SSL self-localization problem while also detecting the environment dynamic objects using onboard monocular vision and inertial odometry data. We enhance the MCL algorithm using insights from implementations of other RoboCup leagues, improving its robustness and increasing its processing speed by adapting the number of particles in the set according to the confidence of the current distribution. For that, we propose a novel approach for measuring the quality of the current distribution based on applying the observation model to the resulting particle of the algorithm. This approach improved up to 56% in computation speed while maintaining the capability to track the robot’s pose.