摘要
当前土地利用边界测量过程中容易受到仪器和人为因素的干扰,导致测量均方根误差(RMSE)较大,为此提出基于全球导航卫星系统(GNSS)技术的土地利用边界测量方法。通过无人机设备和卫星设备对目标土地区域进行微波遥感探测获取大量GNSS反射信号。运用小波分析算法计算GNSS监测信号的一阶导数,从而实现原始信号中的粗差剔除,并对粗差剔除后的信号展开广义延拓插补处理,补充GNSS监测时间序列中的缺失数据。将预处理后的GNSS数据输入构建反向传播(BP)神经网络分类模型中,自动识别出目标区域内的土地利用类型,提取某一类型土地的边界点,完成边界长度和内部面积等土地利用边界的测量。实验结果表明,该方法的土地利用边界测量结果RMSE值不超过0.8,满足土地利用边界的测量要求。
In the current land use boundary measurement process,interference from instruments and human factors can lead to large root mean square errors(RMSE).To address this issue,this paper proposed a land use boundary measurement method based on global navigation satellite system(GNSS)technology.Using unmanned aerial vehicles(UAVs)and satellite equipment for microwave remote sensing detection of the target land area,it obtained a large amount of GNSS reflection signals.The paper applied a wavelet analysis algorithm to calculate the first derivative of the GNSS monitoring signal,thereby eliminating outliers from the original signal.After outlier removal,generalized extrapolation and interpolation processing were performed to supplement missing data in the GNSS monitoring time series.After preprocessing,the GNSS data was input into a back propagation(BP)neural network classification model to automatically identify the land use types within the target area.Boundary points of a specific land type were extracted,and measurements such as boundary length and internal area were completed.The experimental results show that the RMSE value of the land use boundary measurement using this method does not exceed 0.8,meeting the measurement requirements for land use boundaries.
作者
廖伟文
LIAO Weiwen(Zhuhai Surveying and Mapping Institute,Zhuhai,Guangdong 519000,China)
出处
《北京测绘》
2025年第5期689-695,共7页
Beijing Surveying and Mapping
基金
广东省科技计划(2021B1111610001)。
