Analysis of Urban Landscape Change Based on Remote Sensing

Abstract

In this paper, using the multispectral and panchromatic images of the main urban area as experimental data, a method for analyzing dynamic changes of urban landscape based on remote sensing is proposed. The results of landscape elements are obtained through the LVQ2 neural network classification method. From the perspective of patch characteristics, the characteristics of the target urban landscape pattern were analyzed, and the urban change trend was summarized by combining the two-phase images and the transition matrix. The results show that: (1) the number of vegetation in this area has decreased, the area of building land, general land, rivers, and cultivated land has increased, and the greening situation of the city is declining, and the land use has turned to the functional land that improves people's production efficiency; The fragmentation of the landscape patches has decreased, and the connectivity of the landscape has increased. The city government or relevant departments have made overall planning for its overall layout, focusing on the overall distribution of functional areas and various landscapes. The research results of this paper are of great significance for macroscopically grasping the current situation of urban landscape and serving the government and related enterprises and institutions.