Abstract
This paper presents a novel method of seamline determination for remote sensing image mosaicking. A two-level optimization strategy is applied to determine the seamline. Object-level optimization is executed firstly. Background regions (BRs) and obvious regions (ORs) are extracted based on the results of parametric kernel graph cuts (PKGC) segmentation. The global cost map which consists of color difference, a multi-scale morphological gradient (MSMG) constraint, and texture difference is weighted by BRs. Finally, the seamline is determined in the weighted cost from the start point to the end point. Dijkstra’s shortest path algorithm is adopted for pixel-level optimization to determine the positions of seamline. Meanwhile, a new seamline optimization strategy is proposed for image mosaicking with multi-image overlapping regions. The experimental results show the better performance than the conventional method based on mean-shift segmentation. Seamlines based on the proposed method bypass the obvious objects and take less time in execution. This new method is efficient and superior for seamline determination in remote sensing image mosaicking.
Highlights
A large remote sensing image with a wide field of view and high resolution is often required for many applications, such as map-making, disaster management, and military reconnaissance [1,2]
We introduce the method of seamline determination in a two-image overlapping region and can obtain a panoramic image with a wide field of view and high resolution by mosaicking a set of images of 19 using this method
We proposed aa novel based onon segmentation andand the the combined cost for remote sensing image mosaicking
Summary
A large remote sensing image with a wide field of view and high resolution is often required for many applications, such as map-making, disaster management, and military reconnaissance [1,2]. Chon et al [17] used the normalized cross correlation (NCC) to construct a new object function that could effectively evaluate mismatching between two input images This model determined the horizontal expectations of the largest difference in overlapping region, and detected the position of the best seamline using Dijkstra’s algorithm. Pan et al [24] proposed an urban image mosaicking method based on segmentation, which determined preferred regions by the mean-shift (MS) algorithm and calculated the color difference as the cost. Always smaller than the street or the grass, so we cannot accurately extract the preferred regions This method is too simple to construct the cost map, only considering the color difference, but not without obvious objects.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have