北京建筑大学学报2023，Vol.39Issue(6)：72-79,8.DOI:10.19740/j.2096-9872.2023.06.09
基于改进ResNet网络的光伏电站检测方法
Photovoltaic Power Station Detection Method Based on Improved ResNet Network
摘要
Abstract
In order to investigate the negative impact of photovoltaic power stations on the surrounding environment,a method that utilizes remote sensing imagery and deep learning techniques to obtain geographic spatial information of photovoltaic power stations is proposed in the paper.To achieve this,a high-resolution remote sensing imagery dataset for photovoltaic power station object detection is constructed.The multi-level feature extractor structure of the ResNet network is improved,and an error screening-based photovoltaic power station detection method is proposed.The feature extraction capability for photovoltaic power stations in complex and irregular surrounding environments is enhanced and the accuracy of photovoltaic power station object detection is improved.关键词
光伏电站/遥感影像/ResNet网络/目标检测/误差筛选Key words
photovoltaic power station/remote sensing images/ResNet network/target detection/error screening分类
天文与地球科学引用本文复制引用
程洪亮,黄鹤,庞然,杨军星..基于改进ResNet网络的光伏电站检测方法[J].北京建筑大学学报,2023,39(6):72-79,8.基金项目
国家自然科学基金项目(42201483). （42201483）
