![]() It first obtained the mosaic image directly based on seamlines without any blending processing in overlapping areas. Zhu and Qian (2002) presented a hard correction method to remove a possible seam. (2001) presented a feathering method, which used averaging and interpolation functions to eliminate ghosting and reduce intensity differences. They optimized the blending coefficients based on the constraint of the information imbalance between the earlier- and later-acquired images. (2012) improved the pyramid blending approach in for asymmetrically informative biological images during microscope image stitching. Brown and Lowe (2007) also used the same pyramid blending approach in for the panoramic image stitching. Finally, these band-pass mosaic images are summed to obtain the desired image mosaic. Then, different frequency bands were combined with different weighting coefficients and the component images in each spatial frequency band are assembled into a corresponding band-pass mosaic. ![]() Burt and Adelson (1983) presented a pyramid blending approach where images were decomposed into a set of band-pass filtered component images. (2015) improved the weighting function, and presented a cosine distance weighted blending method for high spatial resolution remote sensing images in which the weight calculation algorithm is based on the cosine distance. Milgram (1975) defined a linear ramp to pixel values on either side of the seamline as the weighting function to obtain equal values at the seamline itself. The weighting coefficients vary as a function of the distance from the seamline. Generally, in blending processing, mosaic image I is a weighted combination of the input images I 1 and I 2 over the overlapping areas. Comparisons with other methods further demonstrate the potential of the presented method, as shown in a detailed comparison in three typical cases of the seamline passing by buildings, the seamline passing through buildings, and the seamline passing through areas with large radiometric differences. Experimental results from digital aerial orthoimages in urban areas are provided to verify this method for blending processing based on seamlines in mosaicking. Finally, a multi-resolution reconstruction is performed to obtain the final mosaic. Then, a mask image is generated considering changed regions, and Gaussian and Laplacian pyramids are constructed. The RCR of each region is computed using image segmentation and change detection methods. ![]() The method utilizes the region change rate (RCR) to distinguish changed regions from unchanged regions in overlapping areas. ![]() Therefore, this paper presents a multi-resolution blending method considering changed regions to improve mosaic image quality. Such a mosaic is not a true reflection of the earth’s surface and may have a negative impact on image interpretation. However, for high-resolution aerial orthoimages in urban areas, factors such as projection differences, moving objects, and radiometric differences in overlapping areas may result in ghosting and artifacts or visible shifts in the final mosaic. Blending processing based on seamlines in image mosaicking is a procedure designed to obtain a smooth transition between images along seamlines and make seams invisible in the final mosaic.
0 Comments
Leave a Reply. |