Study On UAV Image Mosaic Algorithm For Territorial Resources Inspection

In recent years,the phenomenon of illegal occupation of territorial resources has occurred from time to time,the regulatory situation for territorial resources is grim.How to monitor and inspect the utilization status of territorial resources in real time is one of the key links for the territorial resources management department to work more reasonably and efficiently.The traditional mode of territorial resources supervision has low efficiency and high cost of manpower and material resources.With the rapid development of Unmanned Aerial Vehicle(UAV)industry,the use of small UAV to collect images has become a new solution to fill the blank of territorial resources supervision.However,as the flying altitude of UAV is low,its single image is small,so it must be stitched together.And sometimes,in order to improve work efficiency or because of environmental restrictions,control points cannot be distributed.Therefore,how to automatically and rapidly splice UAV images without control points has certain research significance.This paper researches and experiments on UAV image stitching technology,and develops a UAV image stitching system without control points for natural resources inspection.The main work includes:(1)The camera parameters are used to correct the geometric distortion of the UAV image to eliminate the radial distortion and tangential distortion.The image enhancement technology is discussed and analyzed in the frequency and spatial domain,and a variety of enhancement technologies are experimentally analyzed.The median filtering technology is selected to eliminate the image noise.(2)Experiments and evaluation analysis of feature point detection were conducted for SIFT algorithm and SURF algorithm.After feature points were detected,Euclidean distance was used as similarity measurement,and BBF algorithm was used as search strategy to roughly match images.Aiming at the problem that there are still a few mismatches after rough matching,RANSAC algorithm is used to delete the mismatches.(3)The linear weighted fusion method was used to conduct the image fusion experiment for the final stitched image of UAV,which eliminated the image patchwork seam and brightness difference.At the same time,because the image mosaics with POS system have low accuracy but the final mosaics have coordinate information,while the image mosaics with feature points have relatively high accuracy but the final mosaics do not have coordinate information,this paper combines the two ideas to complete the image mosaics.The image has coordinate information,which can be superimposed on a variety of spatial base maps,providing convenience for practical applications such as the inspection of natural resources.(4)The image mosaics system developed in this paper was finally combined with other software and hardware devices for operation and experiment.Finally,the obtained images contained coordinate information,and the accuracy could meet the inspection requirements of natural resources.The mosaic images can be used in a variety of practical situations.