DragonFruit-MultiOrgan Dataset

Dataset Description The DragonFruit-MultiOrgan Dataset (DF-MOD) is a comprehensive image dataset containing 5,468 high-resolution images of dragon fruit plants, including fruit and stem/leaf disease conditions. The dataset consists of eight categories: Black Rot, Fungi, Healthy Fruit, Healthy Leaf, Rust, Scab, Stem Canker, and Sunburn. Original Dataset Distribution: - Black_Rot: 675 images (275 MB) - Fungi: 640 images (260 MB) - Healthy_fruit: 650 images (265 MB) - Healthy_leaf: 944 images (305 MB) - Rust: 582 images (187 MB) - Scab: 837 images (292 MB) - Stem_Canker: 553 images (188 MB) - Sunburn: 587 images (189 MB) Total Images: 5,468 Total Size: 1,961 MB The images were collected using an Apple iPhone 13 back dual wide camera with a resolution of 2268 × 4032 pixels, 72 dpi resolution, and 32-bit depth. The camera settings included f/1.6 aperture, 1/577-second exposure, ISO-50, 5 mm focal length equivalent to 26 mm in 35 mm format, automatic white balance, and no flash. Data collection was conducted from multiple ecologically diverse locations in Jhenaidah, Bangladesh, including Charatola, Horinakunda, Krishnapur Road, Baniakandor, Jhenaidah Sadar, Schoolpara, and community-managed agricultural plots. These locations provided natural variations in environmental conditions. All images were carefully reviewed and validated by agricultural experts. The samples were verified based on their disease characteristics, and appropriate class labels were assigned to ensure dataset reliability. Dataset Structure: DragonFruit/ ├── Fruit/ │ ├── Black_Rot/ │ ├── Fungi/ │ └── Healthy/ │ └── Stem/ ├── Healthy/ ├── Rust/ ├── Scab/ ├── Stem_Canker/ └── Sunburn/ A balanced version of the dataset is also provided through augmentation. Each category was increased to 944 images, resulting in a total of 7,552 images. The balanced dataset helps researchers work with uniform class distribution. Balanced Dataset Distribution: - Black_Rot: 944 images - Fungi: 944 images - Healthy_fruit: 944 images - Healthy_leaf: 944 images - Rust: 944 images - Scab: 944 images - Stem_Canker: 944 images - Sunburn: 944 images Total Images After Augmentation: 7,552 The resized dataset and augmented dataset versions are provided separately for researchers based on their requirements. This dataset can be used for various dragon fruit-related research applications, including fruit disease classification, stem disease classification, and combined fruit-stem analysis. The multi-organ structure enables researchers to explore both individual plant-part analysis and innovative approaches involving multiple dragon fruit organs.