A Seamline Optimization Approach Based on Watershed Segmentation for Aerial Image Mosaicking

Shenggu Yuan, Wang Mi, Pan Jun, HU Fen, Dongyang Li · DOAJ (DOAJ: Directory of Open Access Journals) · 2015

Seamline optimization is a key step in the process of aerial image seamless mosaicking.This paper presents a novel algorithm of seamline optimization for aerial image mosaicking by adaptive marker-based watershed segmentation.The preferred region is determined by the difference of the region achieved by adaptive marker-based watershed segmentation. Then, the minimum binary heap Dijkstra's algorithm is adopted to determine the final seamlines in the preferred region. The experimental results show that the seamline determined by our method can avoid crossing obvious stand-alone objects. Compared with other algorithms,our method has higher feasibility and higher speed.

Read the paper · More papers on PaperTik