DEEP LEARNING BASED RAILWAY TRACK INSPECTION

Abstract

Railway infrastructure is crucial for transportation, requiring regular inspection to ensure safety and efficiency. Traditional inspection methods rely on manual labor, which is both time- consuming and prone to human error. This project presents a Deep Learning-Based Railway Track Inspection System that leverages ResNet-based Convolutional Neural Networks (CNNs) to automate defect detection in railway tracks. Developed using Python and Tkinter, the system integrates computer vision and deep learning techniques to analyze railway track images and classify them into different defect categories.

The system features an interactive Tkinter-based GUI, allowing users to upload railway track images, preprocess them into grayscale and binary formats, and analyze them using a trained deep learning model. The classification results indicate whether the track is non-defective, has a fastener defect (requiring immediate replacement or repair), or has a railway defect (demanding urgent maintenance). The model is trained on a dataset of railway track images to enhance its accuracy in real-world defect detection. Additionally, the system includes functionalities such as automated background updates, dynamic title animations, and real-time model training and testing capabilities.

By automating railway track inspection, this project aims to improve safety, efficiency, and accuracy while reducing human effort. The use of deep learning enhances defect detection, making railway maintenance more proactive and effective. This AI-driven approach ensures timely interventions, minimizing the risks associated with track failures and improving overall railway infrastructure reliability.