Analysis of Urban Road Change Characteristics in Gannan Tibetan Autonomous Prefecture Based on Multilayer Neural Network

Roads play a crucial role in urban planning. This study utilizes a deep learning-based Multilayer Neural Network (MNN) to classify road features using Sentinel-2 optical remote sensing imagery from 2019 and 2025. By integrating multispectral characteristics and land cover distribution, the study analyzes the road distribution and its changes in Gannan Tibetan Autonomous Prefecture. The results reveal that road distribution exhibits a spatial pattern of high density in the central urban area and low density in the periphery. Major roads are predominantly radial and grid-like, closely connected to urban building areas, providing convenient transportation support for regional economic development and daily activities.