The development of an intelligent GIS System-A 3D image extraction and construction from a non-digital 2D aerial photograph using digital photogrammetry technique

摘要

The usage of modern appliances in science and technology has enabled us to easily obtain three dimensional (3D) information of target location. However, there is a limitation because past information cannot be retrieved using existing technology. For example, a two dimensional (2D) historical information can be a negative aerial photo, printed aerial photo, map and et cetera. The historical data are used in agricultural mapping, forestry and town planning. A greyscale aerial photo represents information in 2D using x and y values. This research is focussed on height information extraction using image processing technique. In this study, the relationship between pixel and voxel is used to solve high information extraction problem from old aerial photo. Based on Gudes idea, the pixel intensity value of a pixel is related directly to the value of the z-coordinate. Simultaneously, the test on a pixel profile line to obtain the greyscale is implemented. A result of the test is compared using Bersoft Image Measurement, efglabs and Image-Pro Plus. 3Dfield software is used to re-plot the height information extracted from the aerial photo images. Comparisons between the original and re-plotting images have shown insignificant differences. Consequently, it is proven that the proposed technique presented in this report is feasible as a tool for height information extraction from an aerial photo image.