Palora Yellow Pitahayas Dataset with Data Augmentation - Maturity Classification of Pitahaya Fruit

This dataset contains 1,140 RGB images of yellow pitahaya fruit from the Palora variety, classified into three maturity stages: ripe, unripe, and partially ripe. The dataset includes 840 locally acquired images and 300 images collected from publicly available internet sources, with 380 images per maturity class. The images were manually labelled and quality-controlled through duplicate, corrupted, and severely blurred image removal, followed by visual inspection and cross-validation of uncertain labels among multiple labellers. The dataset was resized to 224 x 224 pixels and partitioned into training, validation, and testing subsets. Data augmentation was applied exclusively to the training subset, resulting in a total of 2,736 images used in the experimental protocol, comprising 2,394 augmented training images, 228 validation images, and 114 testing images.