Visual-based trash detection and classification system for smart trash bin robot

This paper presents a trash detection and classification system that will be implemented on a social-education trash bin robot. The robot is expected can be implemented in public facilities, like airport, railway station, hall and more which is there are a lot of people that potentially producing waste. We use Haar-Cascade method to first detect any objects on the floor. Then, Gray-Level Co-Occurrence Matrix (GLCM) and Histogram of Oriented Gradient (HOG) are combined to get a set of features. Support Vector Machines (SVM) is used to classify the features into organic waste, non-organic waste, and non-waste. Offline testing of classification system using 5-fold Cross Validation method obtain 82,7% of accuracy. Online testing of detection and classification system obtain 63.5% of accuracy with the best distance gained when the camera is tilted down to -40° with minimum distance for detection is 80 cm and 200 cm for maximum detection. By using this robot, it is expected to help instill the habit of disposing of garbage in the right place. The purpose of this research is making people aware of handling their waste in the right way and hopefully, it can reduce the waste problem.