基于压缩感知的波形净化及在波形重跟踪中的应用  

Waveform Purification Based on Mixed Gaussian and Its Application in Waveform Retracking

作　　者：陆庆虾 LU Qingxia(Foshan Institute of Surveying,Mapping and Geographic Information,Foshan 528000,China)

机构地区：[1]佛山市测绘地理信息研究院,广东佛山528000

出　　处：《城市勘测》2023年第6期74-77,共4页Urban Geotechnical Investigation & Surveying

摘　　要：受“陆地污染”效应影响,利用卫星测高技术对内陆湖泊水域进行波形重跟踪的精度较低,针对该问题,提出一种基于压缩感知(Compressive Sensing,CS)的波形净化技术用于实现对“陆地污染”引起的异常波峰数据检测及重构,进而结合优化的重跟踪算法进行波形重跟踪改进,提升对内陆水域的测高精度。基于内陆湖泊水文站数据开展验证实验,结果表明,利用净化后的波形进行重跟踪可以明显提升测高精度,相对于传统阈值法和伦敦大学学院法(UCL)精度分别提升28.3%和25.6%,具有较高的应用前景。Affected by the“land pollution”effect,the accuracy of waveform retracking in inland waters using satellite altimetry technology is low.To solve this problem,a waveform purification technology based on mixed Gaussian model is proposed to detect and filter the abnormal peak data caused by“land pollution”,and then combined with the optimized retracking algorithm to improve the waveform retracking and improve the height measurement accuracy for inland waters.The validation experiment is carried out based on the hydrological station data of inland lakes.The results show that using the purified waveform to retrack can significantly improve the height measurement accuracy,which is 28.3%and 25.6%higher than the traditional threshold method and UCL method,respectively,and has a high application prospect.

关 键 词：卫星测高 波形重跟踪 压缩感知 波形净化 内陆湖泊 

分 类 号：P228[天文地球—大地测量学与测量工程]