VIFF: A Visible-Infrared Image Fusion Framework for Enhanced Agricultural Monitoring Towards Agriculture 5.0

Visible and infrared (IR) image fusion plays a critical role in agricultural applications by enhancing image quality and extracting complementary structural and thermal information from different spectral bands. Fusion techniques such as pixel-level, feature-level, and decision-level fusion, aim to generate composite images that retain essential details from both RGB and IR data. However, existing approaches often lack domain-specific adaptation and structured enhancement pipelines suitable for precision agriculture. To address this issue, this work presenting a Visible-Infrared Fusion Framework (VIFF) designed for agricultural monitoring. The framework includes and merges a weighted average fusion algorithm with a series of structural enhancement steps, which include affine transformation, intensity normalization, Gaussian filtering, histogram equalization, unsharp masking, and morphological operations. This method mainly focuses and represents the clarity and maintains important features for crop monitoring and stress detection. Through experiments with apple fruit datasets, the framework shows better image quality and fusion effectiveness, providing a scalable solution that meets agriculture 5.0 goals.