Multi-format open-source sweet orange leaf dataset for disease detection, classification, and analysis.

The Sweet Orange Leaf Disease dataset offers a collection of images of sweet orange leaves collected from sweet orange growing gardens in Khemerdia, Bheramara, Kushtia, Bangladesh, pinpointed at geographical coordinates 35.3602°N and 113.9505°E. This data was collected from August 19 to August 22, 2023. The dataset comprises a total of 5,813 images. The dataset included healthy leaves and diseased leaves. In total, there are eleven classes: Citrus Canker, Citrus Greening, Citrus Mealybugs, Die Back, Foliage Damage, Spiny Whitefly, Powdery Mildew, Shot Hole, Yellow Dragon, Yellow Leaves, and Healthy Leaf. The photo was taken on a sunny day, so the image captured in natural lighting, and the temperature was around 27-29°C. Using two devices explicitly employing the high-resolution camera of an iPhone. The original images, captured in JPEG format, boasted a dimension of 3024 x 4032 pixels. The two devices, the iPhone XS Max and iPhone SE 2nd Generation, captured the dataset's photos. The data are converted into 640*480 pixels and a resolution of 72 dpi. Moreover, the images are annotated based on the diseases. This annotation will assist with semantic segmentation, which has attracted significant attention among data scientists.