| As an important branch of surveying and mapping science, low-altitude photogrammetry takes full digital photogrammetry theory and method as the basis. To a large extent, it makes up for the deficiency of satellites and traditional aerial photogrammetry technology in small areas depending on the advantages of high mobility, low cost, convenience of operation,miniaturization, photography under the low-altitude cloud, customization and so on. As a forceful supplementary method of traditional big-plane aerial photogrammetry, low-altitude photogrammetry increasingly plays an important role in the efficient rapid acquisition and dealing with updating of spatial data. Meanwhile, it has been the hot topic in applied research of photogrammetry domain as well.The paper, on the premise of compliance with relevant laws and standards of unmanned aerial vehicle(UAV), meeting the demand of practical production for purpose, taking UAV regular production process as main line, is researched from the aspects of the framework,integration, big-scale airborne mapping, image stable matching of light-small UAV low-altitude photogrammetry system by looking up plenty of relevant literatures in the domestic and overseas and conducting extensive social research, combining with the accumulated experience of engaging in UAV photogrammetry office and field operation in recent years. The innovation points in this paper can be mainly reflected as the following:(1) Conducting small area and high precision large scale mapping by integration and framework of one UAV low-altitude photogrammetry system.In general, high precision large scale mapping of small area of large rolling terrain depends on total field manual operation mode, as for which, not only the efficiency of mapping is relatively low, but also the cost of men and material is high. Thus, this paper,taking high precision large scale topographic surveying and mapping as the starting points, the required cm-level UAV high-resolution images as the targets, fully analyzes the theory of low-altitude UAV aerial photogrammetry system, and confirms the the key parameters,such as standard flight height, maximum flight height, minimum flight height standard cruising speed and so on. This UAV photogrammetry system uses the relations both between resolution of UAV images and high-precision mapping scale and between theinstantaneous displacement of image point and flight speed. Besides, to configure hardware and architecture with the base-height ratio and heading angle of view, and the relation among the main wheel track, the course and the lateral angle of view. Moreover, it demonstrates the performance of UAV photogrammetry system and that whether it meets the demand of high-resolution large-scale emergency mapping by a series of tests and experiments.(2) Researching the feasibility and measurements of conducting block adjustments using the UAV low-altitude remote sensing images without control points, and the influence of the precision of aerotriangulation using three common adjustment methods.The efficiency of high-resolution mapping is decided directly by the quantity of field control points, especially in the large rolling terrain areas. Reducing, as far as possible, the numbers of field control points will immensely enhance the efficiency of mapping and decline the cost of surveying and mapping on the premise of guaranteeing the precision of mapping. The block adjustments is divided into plane block adjustments and stereo block adjustments in this paper, and demonstrates the conduction of plane block adjustment based on the UAV image with the lack of control points is by using experiment methods. On the contrary, the precision of stereo block adjustments without control points can be improved by augmenting the folds numbers of UAV images acquired in different time in the same area.(3) Researching the tactics of parallel stable matching of low-altitude UAV in the maximum stable areas.Generally, there are problems that the stability of registration and registration accuracy cannot exist simultaneously when conducting images matching with remote sensing stereo images. This paper uses pyramid and partitioning strategies, firstly, to proceed the coarse matching and get a lot of initial feature point set with using the affine transformation to proceed the fitting of maximum stable extremum area in multi-scale space, and then to obtain good initial feature points by limiting the order of the pyramid images to the bad points out and filter homogenization for initial feature points, and finally, to achieve the accurate matching and improve efficiency by conducting parallel SIFT algorithm of image blocking.The matching strategy can effectively improve the efficiency of matching and process the stronger resistance for the rotation,scaling and transformation of the image.(4) Researching the recovery methods of proceeding full-special information usinglow-altitude oblique remote sensing images acquired by using UAV with a tilted camera.Comparing with the traditional modeling method, the modeling of low-altitude images of UAV in the tilt photography way has characteristic of high realism and total factor, but there always exist problems of data loss or detail fuzzy for nearly ground data. In this paper, to overcome the limitations of the titled UVA photogrammetry technology, to couple the titled UAV photogrammetry well with the ground street view technology by improving the construction of tilt camera on airborne platform and the ordinary ground street view, and to recover the loss and fuzzy of details in the model of tilted UAV photogrammetry, then acquiring the tilted model with higher accuracy and recognizable. Then, after comparing and analyzing of restoring after tilt image model and restoring before tilt image model, it is concluded that: the oblique photography coupled with ground street view can immensely improve the authenticity of the data accuracy and detail of images, which close to the ground,and make up for the global limitations and high-level information lost of just using ground street view technology. |
PDF Full Text Request