Quantum Machine Learning

7Development of Sensor-Based Smart Stick with Alert System for Visually Impaired Individuals Using Machine Learning Technique

Sivaramapriya Karuppaiyan¹, Mathivathani Ganesan², Roja Karuppaiyan² and Bhuvaneswari Swaminathan^3*

¹School of Electrical & Electronics Engineering, SASTRA Deemed University, Thanjavur, India

²School of Computing, SASTRA Deemed University, Thanjavur, India

³Department of Computer Science and Engineering, Srinivasa Ramanujan Centre, SASTRA Deemed University, Kumbakonam, India

Abstract

The fusion of machine learning and IoT can potentially drive assistive technologies for visually impaired individuals. However, the existing smart stick works have not failed to integrate hardware and learning models to empower the assistive system. Hence, this proposed work aims to utilize Raspberry Pi, ultrasonic sensors, and a camera module for designing sophisticated IoT, including a Bluetooth module. Meanwhile, an advanced computer vision model has been implemented to achieve high precision in obstacle object detection. For the altered system, the output text from the computer vision model has been given to Google Text-to-Speech, which delivers the audio via a Bluetooth module. Contrary to the previous models that utilized ESP32 microcontrollers without instant obstacle alerts, our smart stick is equipped with a buzzer for safety. These technologies are implicated in one comprehensive and reliable navigation aid, significantly improving the daily activities of visually ...