Multi-Source Pansharpening of Island Sea Areas Based on Hybrid-Scale Regression Optimization

Area of Science:

• Remote Sensing
• Oceanography
• Image Processing

Background:

• High spatial resolution data is crucial for water color inversion in island sea areas.
• Existing multi-source remote sensing data fusion methods face challenges due to sensor biases and spectral distortion.

Purpose of the Study:

• To develop a more precise data fusion solution for enhancing water color inversion accuracy.
• To improve the spatial and spectral quality of fused satellite imagery in complex marine environments.

Main Methods:

• Proposed a novel pansharpening method named Hybrid-Scale Mutual Information (HSMI).
• Integrated mixed-scale information into scale regression for enhanced fusion accuracy.
• Utilized mutual information to quantify spatial-spectral correlations for balanced fusion representation.

Main Results:

• HSMI demonstrated superior performance compared to other popular pansharpening methods.
• The method effectively enhanced spatial details and edge clarity of islands.
• HSMI better preserved the spectral characteristics of surrounding sea areas.

Conclusions:

• HSMI offers a precise solution for multi-source remote sensing data fusion in island sea areas.
• The method significantly improves the accuracy and consistency of water color inversion.
• This approach is valuable for monitoring dynamic marine environments using satellite data.