IEEE Access (Jan 2017)

Locally Linear Detail Injection for Pansharpening

  • Junmin Liu,
  • Yongchang Hui,
  • Peng Zan

DOI
https://doi.org/10.1109/ACCESS.2017.2710226
Journal volume & issue
Vol. 5
pp. 9728 – 9738

Abstract

Read online

This paper presents a novel method, referred to as locally linear detail injection (LLDI), for the pansharpening problem, which is based on the assumption that the spatial details of each multispectral (MS) band can be locally and linearly represented by the spatial details of panchromatic images. LLDI first exploits such assumption through scales using the modulation transfer function (MTF) of the MS instrument and then performs detail injections into the available low-resolution MS images. Visual analysis and quantitative evaluation performed on QuickBird and WorldView-2 data sets at both reduced and full scales show that the proposed LLDI achieves superior improvements over its baselines.

Keywords