Spatio-Temporal Fusion Of UAV Remote Sensing Images Based on Pyramid Method

The current methods used for the fusion of UAV remote sensing images ignore image interpolation, resulting in poor image fusion effect and quality, resulting in large image fusion registration errors. Hence, in this paper, a pyramid-based method is proposed for the spatio-temporal fusion of UAV remote sensing images. Amedian filter is used to denoise the image. The projection transformation model between the two images is described in the form of a matrix and homogeneous coordinate system. According to the respective gray levels between neighboring pixels, the corresponding integer coordinate gray levels in the neighborhood are calculated to perform image interpolation correction, image features are extracted and described, and matching strategies are used for feature matching. The corresponding relationship between features is established, the transformation model is estimated according to the parameters of feature-matching results, and the final transformation matrix is determined to complete the image registration. Through the establishment of a Gaussian pyramid, the Laplacian pyramid is constructed, and the image is reconstructed according to certain rules. By comparing the gray values of the corresponding Laplacian pyramid pixels, the pixels with large gray value are obtained to generate the Laplacian pyramid after fusion, so as to realize the fusion of UAV remote sensing images. The experimental results show that the image fusion effect and the quality of the results obtained by the proposed method are better, and it can effectively reduce the image fusion registration error. © 2022 CRL Publishing Ltd.