Analysis Of 3D Information Reconstruction And Network Optimization Technology Based On Tilt Photogrammetry

From the use of satellites to acquire remote sensing images to small drones to acquire image data,image acquisition methods have become increasingly diverse.Town 3D modeling technology is also becoming more and more mature,but traditional modeling technology has problems such as low efficiency,high cost,poor model authenticity,and low accuracy,which cannot effectively reconstruct large-scale town 3D models.Tilt photogrammetry technology can use the multi-dimensional measurement of flight carrier combined sensors such as drones to obtain image information of a large number of ground objects,which provides a new idea for the efficient production of large-area three-dimensional models.In this paper,oblique photogrammetry technology is the research object,the purpose of the research is to improve the accuracy of the 3D model,and the oblique image matching involved in it is compared and analyzed by three matching algorithms,selecting one of the algorithm as the matching algorithm in this experiment.A mesh model optimization algorithm is proposed.Project the image to calculate the correlation coefficient,and change the triangle shape according to the correlation coefficient to improve the accuracy of the 3D model.The main research contents are:1)Collecting oblique image data in some areas of Taolan County,Gansu Province.By using a fixed-wing drone equipped with a five-lens camera,an aerial photography solution is deployed to obtain the 1770 survey area oblique images and POS data,two representative oblique images were selected for image matching experiments.One group had more trees,fewer houses,and another one was denser.The commonly used principles of the SIFT algorithm,SURF algorithm with affine invariance and ASIFT algorithm with affine invariance are introduced in detail.They are used to perform image matching experiments on two sets of oblique image data.The three algorithms are comprehensively and systematically analyzed in four aspects: correct matching point pair,matching accuracy rate,algorithm operation time,and providing an effective reference for practical production applications.2)According to the conclusions obtained from the image matching experiments,the SIFT algorithm was selected as the oblique image matching method during the 3D reconstruction of the measurement area.The Context Capture software was used to perform the 3D reconstruction of the measurement area,and the modeling process was explained in detail.Software such as Photoshop draws and generates DOM,DEM,DSM,planar elevation renderings,and 3D mesh models to summarize a set of efficient data processing methods.3)This paper investigate the direction of network optimization using 3D modeling based on oblique photogrammetry technology.Based on the 3D mesh model generated by the Context Capture software,network optimization is improved.A mesh model optimization algorithm is proposed and verified by experiments to realize the model and improve accuracy.In the end,three sets of experiments were conducted: a comparative analysis of the areas with fewer buildings,more buildings,and the overall model before and after the optimization of the experimental area.The evaluation was performed from the two aspects of geometric accuracy and model feature performance.The results show that:in the experiment,the optimization effect is better for areas with more buildings and more complicated terrains.The optimized model has outstanding feature performance,and the plane error,elevation error,and three-dimensional error in geometric accuracy have been improved.The model is upgraded after quickly establishing a three-dimensional model accuracy provides new ideas.