Target Classification And Sediment Water Content Estimation In Tidal Flat Using Terrestrial Laser Scanning

Terrestrial Laser Scanning?TLS?has been applied in many fields of earth science due to its accuracy and high resolution.Along with the three-dimensional coordinate information,TLS can simultaneously record the intensity data of each echo that contains physical reflectance characteristics of the scanned point.However,the application of TLS intensity signals is rarely used in tidal flat studies.In this paper,the intensity data is applied to target classification and sediment water content estimation in tidal flat.However,intensity data is affected by the angle of incidence as well as distance effects,and need to be corrected before it can be used.Based on the above reasons,this paper proposes an effective method for intensity data correction,including eliminating the incidence angle effect using indoor target experiments,and the distance effect using homogeneous cement pavement in the field.In combination with the measurement of sediment water content and indoor control experiment,the corrected intensity are used to classify targets in tidal flat of the Yangtze Estuary and predict water content in surface sediments.The results are as following:?1?The intensity correction method proposed in this paper can effectively eliminate the incidence angle and distance effect,and the final corrected intensity data can reflect the surface characteristics of the target directly.In the indoor incident angle correct experiment,a mean reduction of 93%for the coefficient of variation?CV?of the intensity data from four targets is obtained.In the field distance correction experiment,a mean reduction of 83%for the CV of the intensity data from three road pavements is obtained.The results indicate the correction methods is effective and applicable to different target.?2?The intensity correction method is used to correct the point cloud data of tidal flat on the Chongming Island,Shanghai.Then the corrected intensity is used to classify the targets in tidal flat,including bare flat,vegetation and concrete road.The CV of the corrected intensity data is reduced by 54%compared to the original intensity data.Using the k-means clustering algorithm,theF₁-Score indicator in the confusion matrix indicates that the classification accuracy of final corrected intensity is 81%.In comparison with the original intensity data,the classification accuracy is improved by49%.?3?The indoor experiment show that exponential function can be used to fit the relationship between sediment water content and corrected intensity.Applying this function to surface sediments in tidal flat on the Chongming Island,Shanghai,the root-mean-square error between the measured and predicted water content is 3%?4?This study demonstrates that the corrected intensity data can be effectively used for target classification and sediment water content estimation.