Vision Based Pay-as-you-throw system: motion detection for garbage and recycle - based on faster RCNN

Loading...
Thumbnail Image

Keywords

Degree Level

Advisor

Degree Name

Volume

Issue

Publisher

Memorial University of Newfoundland

Abstract

This project implemented the motion detection and tracking, the goal of project is detection, tracking and counting the different types garbage. The Faster Region Convolutional Neural Network (Faster RCNN or FRCNN) is used as classi�er and detector. Because the count for garbage or recycle is the �nal result, therefore, the tracker is developed based Kalman �filter