研究目的
Evaluating the performance of full waveform LiDAR decomposition algorithms with a high-resolution single band airborne LiDAR bathymetry system in shallow rivers, and proposing a continuous wavelet transformation method for waveform processing.
研究成果
Full waveform processing, particularly using CWT, provides denser point clouds and better discrimination of benthic returns compared to discrete methods. However, no single algorithm is superior for all conditions; performance depends on water turbidity and surface definition. Optimal strategies achieved water depth accuracies of 6 cm mean and 14 cm standard deviation in clear water, and 16 cm mean and 27 cm standard deviation in turbid water. Multi-wavelength systems may be necessary for highly turbid environments.
研究不足
The study is limited to shallow river environments with specific turbidity conditions; performance may vary with different water clarities and depths. The single-band LiDAR system may not perform as well as multi-band systems in turbid water. The methods require accurate initial peak estimates and are computationally intensive. Future work is needed to improve water surface extraction in complex backscatter conditions.
1：Experimental Design and Method Selection:
The study involves comparing different full waveform processing algorithms, including continuous wavelet transformation (CWT), Gaussian decomposition, and empirical system response (ESR) methods, applied to airborne LiDAR data for bathymetric mapping in shallow rivers.
2：Sample Selection and Data Sources:
Two datasets from the Snake River and Blue/Colorado River were used, with varying water turbidity. Data collected using Optech Aquarius and Gemini LiDAR systems, and ground truth from Acoustic Doppler Current Profiler (ADCP) and Real-Time Kinematic (RTK) GPS measurements.
3：List of Experimental Equipment and Materials:
Optech Aquarius LiDAR system, Optech Gemini LiDAR system, Sontek RiverSurveyor S5 ADCP, WET Labs EcoTriplet turbidity sensor, RTK GPS, and TerraScan software.
4：Experimental Procedures and Operational Workflow:
Pre-processing of LiDAR waveforms by noise thresholding, application of CWT, Gaussian decomposition, and ESR methods to extract peak locations and generate point clouds. Classification of benthic and water surface points, calculation of water depths using point-to-plane distance, and comparison with ADCP measurements.
5：Data Analysis Methods:
Statistical analysis including mean bias, standard deviation, slope, intercept, and R-squared values for depth comparisons; use of least squares estimation and regression analysis.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
