DETECTING HIDDEN DANGERS IN RAILWAY CONSTRUCTION USING MACHINE VISION TECHNOLOGY

Abstract

In order to achieve accurate identification of safety hazards at railway construction sites, taking a railway engineering project as an example, machine vision technology was introduced to design a method for identifying safety hazards at railway construction sites. In this study, the images of key monitoring areas are first collected, input into the computer, and preprocessed. Then, the scene model of the construction site is constructed through the camera and infrared thermal imaging acquisition equipment, and the image background model of the dangerous area is established by referring to the Gaussian model. Finally, the image carrying the background is input, and the image is matched with the time series. The identification conditions of construction site dangerous areas are set up to realize the identification of hidden danger points in the background of dangerous point image. Comparative experimental results show that this method can improve the identification rate of safety hazards in railway construction sites.