This technology is a remote image processing method and device that generates high-resolution multisensor images by decomposing satellite multispectral (MS) and panchromatic (PAN) images into high- and low-frequency components using a guided filter to create an initial detail map, enhancing it with spectral information extracted from the MS image, and applying adaptive coefficients.
Due to the limitations of satellite optical sensors, it has been difficult to perform precise analysis because only low-spatial-resolution MS images and low-spectral-resolution PAN images could be obtained. Existing multiresolution analysis methods require complex decomposition tools and cause spatial distortions such as ringing and local blurring, while component substitution methods and IHS-based techniques suffer from significant spectral distortion even when spatial resolution is improved.
This technology balances spectral and spatial information by incorporating a two-stage residual into the initial detail map to efficiently reinforce spectral information and refining the spatial structure using adaptive coefficients from the AIHS technique. It can be applied to satellite image analysis services for disaster prediction, environmental monitoring, and urban observation, allowing for high-resolution, color-accurate imagery even from low-cost satellite data.
N/A
