AI-Powered Distance Estimation for Autonomous Systems: A Monocular Vision Approach

Abstract

Accurate distance estimation is crucial for autonomous systems to navigate safely in dynamic environments. This study explores a novel deep learning-based approach to estimate object distances using monocular vision, eliminating the need for costly LiDAR sensors. By leveraging supervised neural networks and representation-based regression models, I improve real-time depth perception for self-driving applications. The methodology includes model adaptation, prototype development, and comparative performance evaluation across multiple architectures. Experimental results demonstrate the feasibility of low-cost, high-accuracy distance prediction, paving the way for enhanced situational awareness in autonomous vehicles.