View Proposal

---

Proposer

Chee Een Yap

Title

SmartBin: An IOT Real-Time Waste Classification and Segregation System with Zero-Shot learning

Goal

To design and implement a low-cost, real-time autonomous waste classification and segregation system based on Raspberry Pi, camera vision, and machine learning algorithms for accurate waste recognition, reduced manual sorting effort, and enhanced environmental protection.

Description

The proposed system is an autonomous waste classification and segregation system consisting of a Raspberry Pi, camera module, infrared sensor, waste bins, LED indicator lights, and electromagnetic door locks. The camera, Raspberry Pi, and infrared sensor are installed between the two waste bins to detect and classify waste. When a waste item is placed in front of the system, the infrared sensor detects it and captures an image. The captured image is then processed locally by the Raspberry Pi with Zero-Shot Learning used to classify the waste items. Zero-Shot Learning is expected to recognize the unseen waste items without extensive training data. Based on the classification result, the system automatically performs waste segregation by: 1. Unlocking the corresponding waste-bin door using an electromagnetic lock. 2. Turning on the LED indicator light associated with the selected bin. 3. Guiding users to dispose of the waste into the correct bin. The entire process is performed locally on the Raspberry Pi without requiring continuous cloud connectivity, ensuring low operational cost, enhanced privacy, and reliable deployment in environments with limited internet access.

Resources

Background

Url

Difficulty Level

Moderate

Ethical Approval

None

Number Of Students

1

Supervisor

Chee Een Yap

Keywords

raspberry pi, zero-shot learning, waste classification

Degrees

Bachelor of Science in Computing Science