Vision Navigator: A Smart and Intelligent Obstacle Recognition Model for Visually Impaired Users

Table 1

Hardware and software components of vision navigator.

System Requirements	Description
Arduino board	Controls, processes, and generates all inputs and outputs. It receives the echo signals from the ultrasonic sensor that trigger it to take further actions and checks if the obstacle is there. It generates an immediate alert using a buzzer. It also generates a caption for the image captured by a camera and later converts that caption into speech that is played through an audio device.
Ultrasonic sensor	Determines the target obstacle distance by emission of ultrasonic sound waves, which maps the reflected sound into an electrical signal.
IR sensor	Detects infrared radiation that helps in sensing obstacles in surroundings.
Vibrator	Used to notify the user about any obstacle present.
Bluetooth module	Used to send and receive the signal through two devices in a wireless medium.
Push buttons	Used to switch on the microcontroller board.
Water sensor	Used to notify the user of any water body presence.
Audio module	Conveys image captions to the user in the form of audio. It receives an audio signal from Arduino once the caption for the image is converted into an audio format using a text-to-to-speech algorithm.
Camera modules	Acts as eyes for visually impaired people. Each time an ultrasonic sensor detects an obstacle, the camera modules capture the picture that is sent to the board for processing and caption generation.
Buzzers	Used in this system for an immediate alert. When the ultrasonic sensor detects an obstacle, it is triggered.
Plastic stick bodies	Acts as the outer texture body of the Smart-fold Cane.
Python	Programming interface to implement the model.
TensorFlow	Open-source machine learning platform.
Text-to-speech API	Application to map the obstacle text into speech to notify users.