Research On UAV Data Fusion For Characteristic Sparse Waters

With the continuous development of UAV technology,UAV has been applied to topographic survey,precision agriculture,disaster relief and many other fields.Compared with satellite remote sensing technology,UAV technology has the advantages of low cost,short cycle and high precision.Therefore,using UAV to collect images of a certain area has become a more common monitoring and mapping method.The mapping task of UAV needs to splicing aerial images to get the panoramic image of the area to be mapped.However,the current aerial image Mosaic methods are mostly based on image features.These methods complete image registration and image Mosaic by extracting the set of common feature points between images.The method based on image features can work effectively for aerial images with rich features,but this method will fail to find enough common features for aerial images with sparse features.In order to complete the mapping task in the special situation of sparse water areas,the following aspects are studied in this paper:(1)In view of the problem that the image feature-based Mosaic method cannot work in sparse feature waters and the Mosaic speed is slow,a method of image rapid Mosaic using the position and attitude information of aerial photography is proposed.Firstly,the optimized aerial pose data is obtained from the flight log of UAV equipped with GPS and IMU.Then the transformation matrix is calculated according to the position and attitude data of aerial images and all images are converted to a unified projection coordinate system.The image is registered according to GPS data.Finally,image enhancement and image fusion algorithm are used for the Mosaic image.The experimental results show that the proposed method is better and faster than the traditional feature-based method.(2)In order to improve the quality of aerial image Mosaic and improve the visual effect of the image,this paper preprocesses the original image through the method of adaptive histogram equalization to achieve the purpose of image enhancement.Experimental results show that this method can effectively improve the information entropy and average gradient of the image,and the image quality can be effectively improved.In order to solve the problems of exposure difference and stitching error in aerial image stitching,image fusion method is used to make the overlapping area transition naturally.In this paper,several image fusion methods are compared and their advantages and disadvantages are discussed according to the experimental results.(3)Considering the complexity of the mapping task,the UAV mapping system for characteristic sparse water area is designed and implemented.This paper puts forward requirements for UAV hardware based on surveying and mapping tasks,and uses related software to complete the surveying and mapping tasks of UAV.Firstly,the automatic path planning and the automatic flight of UAV are carried out by the task planning program.Then the sensor data and images are processed by the data processing and aerial image Mosaic software implemented in this paper to obtain a panoramic image containing geographic information.Finally,GIS application is used to display and analyze aerial images.The experiment proves that the UAV mapping system designed in this paper can work effectively in characteristic sparse water area.