SoilWet-5: Image Dataset of Sandy Loam Soil with Five Wetness Levels from the Jamuna Riverside of Tangail District, Bangladesh

SoilWet-5 is an image dataset of sandy loam soil collected from various land sources along the Jamuna riverside in Tangail District, Bangladesh. The dataset contains 750 images equally distributed across five wetness levels: Very Wet, Wet, Moisture, Dry, and Very Dry, with 150 images in each category. It was developed to support research in soil wetness classification, computer vision, image processing, machine learning, deep learning, and explainable AI-based soil analysis. The dataset may be used for agricultural monitoring, texture analysis, feature extraction, and intelligent soil condition recognition.