Spatial and Temporal Evolution of Vegetation Based on Optical Flow Algorithms

Vegetation plays a critical role in ecological studies, and the identification of vegetation is a key focus of remote sensing vegetation monitoring, which is essential for applying remote sensing to the study of ecological environments. In order to accurately identify the vegetation, dynamically monitor changes in vegetation cover, and analyze its spatial and temporal evolution, this study uses Normalized Difference Vegetation Index (NDVI) for vegetation identification, and uses Convolutional Neural Networks (CNNs) combined with optical flow algorithms to visualize and analyze the identification results, which intuitively demonstrates the pattern of change of vegetation in different years. The results are as follows: The FlowNet2.0 model enables clear visualization of vegetation changes, and the feasibility of employing the optical flow algorithm for visualization purposes is confirmed through the analysis of vegetation increase and decrease status. This paper proposes a new method that combines CNNs and optical flow algorithm for visualizing vegetation changes.