Improved Performance of Trash Detection and Human Target Detection Systems using Robot Operating System (ROS)

In a visual-based real detection system using computer vision, the most important thing that must be considered is computation time. general, has heavy algorithm puts strain on performance of system, especially if to handle two or more different processes. This paper presents an effort improve trash and target partner bin robot with social interaction capabilities. The uses combination Haar Cascade algorithm, Histogram Oriented Gradient (HOG) Gray-Level Coocurrence Matrix (GLCM). Meanwhile, Depth algorithms. Robotic Operating System (ROS) used make each in separate modules which aim utilize all available resources while reducing As result, obtained by ROS platform capable running at speed 7.003 fps. human 8,515 line increase fps, accuracy also increases 77%, precision 87,80%, recall 82,75%, F1-score 85,20% detection, improved 81%, %, 91,46%, 86,20%, 88,42%.