Multi-Slice Image Mapping Research Based On UAV Oblique Images

With the continuous development of digital photogrammetry technology,it has gradually become an important way to acquire and update national basic geographic information data.It provides a rich and varied information source for the measurement and updating of national geographic information data.At the same time,digital photogrammetry technology also plays a key role in the construction of national basic economy.It is widely used in the fields of natural disaster monitoring and prevention,research of land resources,digital city construction and management,urban planning and design.UAV oblique images technology has been the key technology of digital photogrammetry,and it is also the key content of joint research in digital photogrammetry and computer vision.Multi-slice images mapping technology based on oblique images makes full use of the advantages of high overlap and multi-angles of oblique images,selects multiple image pairs with better base height ratio,and simultaneously cuts through a feature with the same name point in multiple images.According to the extra observation,the high-precision object-side measurement coordinates are obtained by using the least square iterative solution to get a better accuracy of the object coordinates.In this paper,based on UAV oblique images of multi-slice images mapping were studied and discussed,the study summarized UAV oblique image multi-slice images mapping methods and principles,focusing on UAV oblique multi-slice image mapping and how the two major aspects of the oblique image with good angle of view are selected for the algorithm research and demonstration experiment.Aiming at the problems of non-observational redundancy and small base-height ratio,poor accuracy and poor robustness caused by the traditional cross-sectional mapping method,a multi-slice images measurement based on oblique images is improved,this method is not easily limited by the size of the intersection angle,the use of multi-slice images forward intersection,while the introduction of least squares principle,you can get more accurate coordinates of the object.As to how to obtain a good angle of view of the oblique images,due to the UAV when shooting,the heading of overlap shall not be less than 60%,the degree of overlap shall notbe less than 30%.Therefore,there are a large number of repeated observations.In this paper,six methods are selected to screen the oblique images,such as light direction screening,object range screening,intersection angle screening,projection screening,distance screening with image center,and resolution.Finally,the effectiveness of this method is verified by experimental analysis.Aiming at multi-slice images of UAV oblique images,this paper adopts two sets of experiments and demonstrates the image data obtained from the three regions.The final experiment shows that:(1)The errors of the multi-slice images mapping method described in this paper is smaller the three-dimensional mapping.(2)When the image sharpness is good,the accuracy of the coordinates measured by the multi-slice images mapping is better than the measured by RTK.Similarly,when the UAV's oblique image sharpness is normal,the accuracy of the coordinates measured by the multi-slice images mapping is compared with the measured by RTK,and the accuracy is also good.At the same time,the overall accuracy of the calibration is also relatively good.