A Multi-Resolution Blending Considering Changed Regions for Orthoimage Mosaicking

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. 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. Such a mosaic is not a true reflection of the earth’s surface and may have a negative impact on image interpretation. Therefore, this paper presents a multi-resolution blending method considering changed regions to improve mosaic image quality. The method utilizes the region change rate (RCR) to distinguish changed regions from unchanged regions in overlapping areas. The RCR of each region is computed using image segmentation and change detection methods. Then, a mask image is generated considering changed regions, and Gaussian and Laplacian pyramids are constructed. Finally, a multi-resolution reconstruction is performed to obtain the final mosaic. Experimental results from digital aerial orthoimages in urban areas are provided to verify this method for blending processing based on seamlines in mosaicking. 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.

[1]  David L. Milgram,et al.  Computer Methods for Creating Photomosaics , 1975, IEEE Transactions on Computers.

[2]  Chun-Shin Lin,et al.  Seam-line determination for image mosaicking: A technique minimizing the maximum local mismatch and the global cost , 2010 .

[3]  Matthew A. Brown,et al.  Automatic Panoramic Image Stitching using Invariant Features , 2007, International Journal of Computer Vision.

[4]  Huanfeng Shen,et al.  A robust mosaicking procedure for high spatial resolution remote sensing images , 2015 .

[5]  Yuan Yao,et al.  Geomatics for Smart Cities - Concept, Key Techniques,and Applications , 2013, Geo spatial Inf. Sci..

[6]  Chi-Keung Tang,et al.  Image Stitching Using Structure Deformation , 2008, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[7]  Yizong Cheng,et al.  Mean Shift, Mode Seeking, and Clustering , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  D. Lu,et al.  Change detection techniques , 2004 .

[9]  Ashbindu Singh,et al.  Review Article Digital change detection techniques using remotely-sensed data , 1989 .

[10]  Edward H. Adelson,et al.  A multiresolution spline with application to image mosaics , 1983, TOGS.

[11]  Wen-Liang Hwang,et al.  Optimal Multiresolution Blending of Confocal Microscope Images , 2012, IEEE Transactions on Biomedical Engineering.

[12]  Martin Kerschner,et al.  Seamline detection in colour orthoimage mosaicking by use of twin snakes , 2001 .

[13]  Shmuel Peleg,et al.  Seamless image stitching by minimizing false edges , 2006, IEEE Transactions on Image Processing.

[14]  Dorin Comaniciu,et al.  Mean Shift: A Robust Approach Toward Feature Space Analysis , 2002, IEEE Trans. Pattern Anal. Mach. Intell..