Compression of Three-Dimensional Terrain Data Using Lifting Scheme Based on Second Generation Wavelets

The most complex GIS data are three dimensional terrain data. In GIS applications, for a realistic representation of a terrain the Digital Elevation Model (DEM) is not suitable for direct use in online GIS services due to its large size and inflexibility data structure. Triangulated Irregular Network (TIN), another popular data format for three dimensional data, requires great number of triangles. These triangles that represent the surface of the terrain ultimately increase the data size. For online GIS interactive programs it has become highly essential to reduce the number of triangles in order to save storage space. Existing image compression systems for GIS terrain data have bandwidth and image size constraints that result in time-consuming transmission of uncompressed raw terrain data. Thus image compression is a key factor to improve transmission speed and storage, but it risks losing relevant terrain information. Moreover, most of these algorithms are based either on Fourier transmission or on first order wavelet techniques. Very little work has been done till date on the GIS terrain data compression based on second generation wavelets. Second generation wavelet technology provides an efficient compression tool to achieve high compression ratio while maintaining an acceptable fidelity of surface quality. The primary motivation for this work stems from the fact that there is need for a new spatial data compression technique for GIS data compression. This thesis presents a new data compression technique using lifting scheme based on second generation wavelets. The lifting scheme has been found to be a flexible method for constructing scalar wavelets with desirable properties. In this thesis, it is extended to the GIS data compression. A newly developed data compression approach to approximate the terrain surface with a series of non-overlapping triangles has been presented. Generally a Triangulated Irregular Networks (TIN) is the most common form of digital surface model that consists of elevation values (z) with x, y coordinates that make up triangles. Firstly, the irregular sets of points were taken and they are used it to find average signal and difference signal (detail coefficients). Delaunay triangulation and bivariate splines are used to estimate average signal and difference signal. This approach covers following steps: First, by using a Delaunay triangulation, a TIN representation of the terrain from an arbitrary set of data is generated. A new interpolation wavelet filter for TIN has been applied in two steps, namely splitting and elevation. In the splitting step, a triangle has been divided into several sub-triangles and the elevation step has been used to ‘modify’ the point values (point coordinates for geometry) after the splitting. Then, this data set is compressed at the desired locations by using second generation wavelets. High difference signal or detail coefficient value indicates significance of a point. Only the set of significant points are used to represent the terrain. The bivariate splines are used to quantize the signal over Delaunay triangulation. The size of this set of significant points will become very small compared to the original data set and hence the data file will be compressed. This data set can be transferred easily and terrain image can be regenerated by using a program based on Delaunay triangulation and bivariate splines. The image processing toolbox of the MATLAB (version 7) is used to develop programs based on the lifting scheme for multiresolution representation of terrain. The Mean Square Error (MSE) and Peak-Signal-to-Noise Ratio (PSNR) are calculated. The newly developed algorithm was applied to compress Light Detection and Ranging (LIDAR) data to check the efficiency of the program. The quality of geographical surface representation after using proposed technique is compared with the original LIDAR data. The results show that this method can be used for significant reduction of data set.

Preview PDF 1600459.pdf Download (112kB)

Actions (login required)

View Item