Implementation of 2D Lidar Sensors Using Maximus Tracking Clustering Algorithm and Natural Landmark Using AGV Rover Robot

This paper introduces the Maximus Tracking Algorithm, which is a mixed method that uses Euclidean clustering to improve edge line mapping and DBSCAN for identifying objects, helping to clearly tell the difference between trees and walls in outdoor areas. The algorithm uses data from a depth camera just to check how accurate the mapping is, not for finding objects itself. By doing this, the system improves SLAM performance by allowing real-time filtering of point cloud data and better matching of features, leading to more reliable and precise 2D maps for AGV rovers. Field tests show the system works well, with mapping accuracy reaching up to 98%. Comparing results with and without the algorithm shows notable improvements in map clarity, edge detection, and stable navigation. Statistical analysis from several test runs shows consistent performance with a standard deviation of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\pm 1$</tex>. 2<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>, showing the method is dependable.