Multispectral band sharpening using pseudoinverse estimation and fuzzy reasoning

Abstract

Exploitation of commercial multispectral satellite imagery (e.g., Landsat Thematic Mapper (TM) and SPOT multispectral scanner) can be extremely useful for surveillance and broad area search due to the large geographic areas covered with each orbital pass. However, most multispectral sensors are not designed for the specialized tasks associated with surveillance. As a result, multispectral exploitation for surveillance faces significant technical problems. Principle among these problems is the low spatial resolution of the sensor. This paper presents an innovative technique that synergistically fuses high resolution panchromatic imagery with lower resolution multispectral imagery to generate a "sharpened multispectral image", which is an estimation of high resolution multispectral information. This is a two stage paradigm where an initial estimate of the sharpened image is made using the pseudoinverse and then refined using a set of fuzzy rules. The pseudoinverse produces a minimum mean-squared error estimate of the sharpened pixel while the fuzzy rules refine this estimate using the local information contained in the surrounding pixels. This technique and preliminary processing results are presented. Implications for surveillance is also discussed.

Real-GDSR: Real-World Guided DSM Super-Resolution via Edge-Enhancing Residual Network

“…The idea is that a different imaging modality of the same object can be used as a guide for super-resolving the low-resolution image by injecting the missing high-frequency content. Research into GDSR has a long history (Patterson et al, 1992;Izraelevitz, 1994). The proposed solutions range from classical, entirely hand-crafted schemes (Ham et al, 2017) to fully learning-based methods (Hui et al, 2016a), while some recent works have combined the two with promising results (Lutio et al, 2022;Metzger et al, 2023).…”