A Smartphone-Based Computer Vision Assistance System with Neural Network Depth Estimation for the Visually Impaired

Abstract

We propose a smartphone-based computer vision system for visually impaired people that uses a neural network to classify objects and estimate image depth to improve spatial orientation in the environment. For this purpose, we have developed and implemented a spatial orientation algorithm with a recursive function for calculating the sum of image array values to estimate depth. The advantage of this algorithm is the low complexity of calculations, which ensures its high performance in real-time. Our system is designed to be easy to use, portable, and affordable, making it accessible to a wide range of users. The proposed system utilizes a smartphone camera and computer vision algorithms to analyze the user’s environment and provide real-time feedback through audio and haptic feedback. The neural network depth estimation model is trained on a large dataset of images and corresponding depth maps, which allows it to accurately avoid various objects in the user’s field of view.