Design of an intelligent robotic donation box a case study

Abstract

The design and development of an Intelligent Robotic Donation Box (IRDB) system as a typical final year project in Mechatronics engineering is presented in this paper. The developed IRDB system has the capability of collecting donation from people in an organized assembly or gathering. Also incorporated into this system is the ability to attract attention of people by making audible sound, gesture recognition as well as to avoid any obstacles on its path. New skin detection algorithms developed from artificial neural network and pseudo-modeling technique using YCbCr color spaces have also been proposed in this paper. Performance evaluation of the proposed new skin extraction techniques shows very promising results with an average accuracy of 92.75% and false positive rate of 4.5%. Similarly, performance evaluation of the IRDB system shows that the system can conveniently replace the existing manual boxes used in some organized gathering around the world and yet at a reasonable cost as compared to the existing donation boxes.