Obstacle Detection and Tracking System for Robot Soccer ERSOW based on Centroid Tracking Method

摘要

EEPIS Robot Soccer on Wheeled (ERSOW) is a wheeled soccer robot developed by Politeknik Elektronika Negeri Surabaya and participated in the national robot competition, Indonesian Robot Contest in KRSBI Wheeled division. ERSOW team has striker robot, striker robot was used in attack strategy for scoring goals. The ability to tracking the obstacle is one of main abilities that striker robot must has. Obstacle tracking plays a crucial role in robot soccer on wheels due to the dynamic and unpredictable nature of the game. As robots navigate the field, they encounter various obstacles such as opponents, teammates, and static objects (obstacle). Accurate and real-time tracking of these obstacles enables robots to make informed decisions, adapt their strategies, and effectively interact with the environment. Obstacle tracking enhances situational awareness. By continuously monitoring the positions and movements of obstacles, robots gain valuable information about the current game state. Tracking system on this research works with centroid tracking algorithm. The fundamental concept of the centroid tracking algorithm revolves around assigning a distinct identification (ID) to every tracked object by taking their centroids, this algorithm takes the center of obstacle bounding box as an input, to generated bounding box this system used thresholding detection method based on HSV color space. The center of bounding box (in pixel coordinate) converted into world coordinate by using Ab-exponential regression and Linear regression. The testing results of tracking system could estimate obstacle from 50cm–250cm from the robot with 0,967 cm error average, the system also could track the obstacle with 86% accuracy when the robot moved 330 cm ahead.