LOW-CARBON VEGETATION CONFIGURATION STRATEGIES FOR URBAN PARKS USING DEEP LEARNING TECHNOLOGY

Applied Ecology and Environmental Research 2026 Open access

Read full text
Authors: Y. AN
DOI: 10.15666/aeer/2401_199215
Cited by: 0
References: 21

Abstract

This study utilizes the DeepLabV3+ panoramic segmentation model based on ResNet50 to identify and quantitatively analyze the landscape elements in urban parks in China.The aim is to propose low-carbon vegetation configuration strategies to optimize park landscape design.By obtaining pictures of park landscape scenes and pre-training the model with the public ADE20K dataset, high-precision segmentation of landscape elements is achieved.The study quantitatively characterizes the landscape elements and focuses on analyzing the correlation between the Green Visibility Rate (GVI) and the vegetation proportion, as well as their impact on visitor satisfaction.

Fields of Study & Subjects

Urban Arborization and Environmental Studies Urban Green Space and Health Urban Planning and Landscape Design

Publication details

Journal / source
Applied Ecology and Environmental Research
Volume
24
Issue
1
Pages
199–215
Publication date
1 January 2026
Language
EN