Pineapple Leaf Disease Dataset: Field-Acquired Images for Deep Learning-Based Plant Health Assessment

This dataset contains original field-acquired images of pineapple leaves for deep learning-based disease classification and precision agriculture research. Images were collected from pineapple cultivation regions in Vazhakulam, Kerala, India. The dataset consists of four classes: Healthy, Mealybug Wilt, Fusarium Rot, and Leaf Blight. For each class, 50 unique pineapple leaf specimens were selected, and each specimen was photographed with a minimum of 20 distinct original images from different viewpoints, distances, and orientations. The dataset contains a total of 4,476 original images, comprising 1,157 Fusarium images, 1,118 Healthy Leaf images, 1,112 Leaf Blight images, and 1,089 Mealybug Wilt images. All images were captured under natural field conditions with the leaves remaining attached to the pineapple plants; no leaves were cut or detached for image acquisition. The dataset contains original images only and does not include artificially augmented images. Each specimen is assigned a unique ID, and the images are organized class-wise and specimen-wise. Metadata describing the image name, class, specimen ID, collection information, camera is provided with the dataset. The dataset is intended for research in pineapple leaf disease classification, computer vision, deep learning, and precision agriculture.