Research On The Method Of Generating DEM With LiDAR And Multisource Data

Integrated with GPS, inertial navigation and laser ranging technology, LiDAR(Light Detection And Ranging), as an active geo-measurement system, has such characteristics as little dependent on ground controls, little affection by weather conditions, high automatic and short data processing temporal, and it would probably bring a new technical revolution for surveying, and consequently gives geographers a new challenge(Liu Jingnan, 2003).LiDAR can quickly get accurate and high resolution DTM and three-dimensional coordinate of ground feature. Due to its rapid acquiring speed, high vertical accuracy and low processing cost, as well as the advantages of weather-independent, coast-survey and bathymetric surveying, LiDAR fits greatly to the generation of DEM. Therefore, Airborne LiDAR system has gradually become one of the most advanced remote sensing systems, and it would have wide development foreground and application demand in the field of Photogrammetry and Remote Sensing.In current commercial laser measurement system, feature extraction, classifying, and DEM generating for LiDAR data need lots of artificial assisted operating. High accuracy three-dimensional DTM can be obtained by filtering and interpolating LiDAR data logically. So far, great researches have been done on such matter, and as a result, many filtering and classifying algorithms have been put forward. However, these filtering algorithms are non-intelligent image processing methods in which the ground feature laser points which need to be reserved and the target points which need to be filtered can not be distinguished. As a result, the DEM obtained from LiDAR data is generally smooth, and some important ground feature information may be lost.This paper mainly studies the theory and method of filtering LiDAR data. On the basis of analyzing and concluding the theory and characteristics of LiDAR, this paper discusses the key technology of LiDAR data, and advances the thought of generating DEM with LiDAR and multi-source data. The specific contents are as follows:1. Present situation of generating DEM with Airborne LiDAR survey and photogrammetry, and a comparison between the two technologies;2. Discussion on the characteristics and organization of the LiDAR data, standard file format proposed by ASPRS, which is a great significance for system providers software developers and users;3. The research on quality evaluation and gross error analysis methods for LiDAR data; classification and conclusion of airborne LiDAR systematic error; proposing a gross error analysis method based on area of strips overlap;4. Evaluating the current filtering algorithms synthetically; advancing the method of generating DEM with LiDAR and multi-source data; discussing the theory of registering multi data and filtering algorithm; filtering algorithm experiment with LiDAR data of mountainous region and a preferable conclusion.5. Advancing a DEM accuracy evaluation method based on terrain slope; getting a useful conclusion by calculating grid of two kinds of DEM in different slope extents by using grid calculator and map algebra language.