DeepSort-3C Background-Removed Object Dataset for Intelligent Conveyor Belt Classification

This dataset, titled “DeepSort-3C Background-Removed Object Dataset for Intelligent Conveyor Belt Classification”, is developed for industrial object detection and classification in automated conveyor belt systems. The primary objective of this work is to enhance real-time object tracking and classification performance using clean, background-removed images. The dataset consists of carefully processed and annotated images categorized into three distinct classes: colorful objects, transparent objects, and black objects. These classes were specifically designed to support robust classification in varying visual conditions commonly found in industrial environments. Background removal techniques were applied to reduce noise and improve model accuracy for deep learning-based systems. Each object class was prepared and labeled to ensure proper classification for training machine learning and deep learning models. The dataset supports improved object recognition in scenarios where traditional datasets fail due to complex backgrounds and low visual contrast. This project integrates DeepSort-based object tracking concepts along with dataset preparation techniques to support intelligent industrial automation. It aims to address challenges such as cluttered environments, limited labeled data, and inefficient object recognition in real-world manufacturing systems. The dataset can be utilized for research and development in the fields of computer vision, deep learning, object detection, and industrial automation systems.