Comprehensive lychee leaf data set: Digital images for analysis and classification

Published: 26 September 2023| Version 1 | DOI: 10.17632/xf7ryhs3cd.1


The lychee leaves in the investigated dataset were carefully selected for a variety of research and analysis uses. The 4,129 high-resolution images in this dataset each show intricate lychee foliage details. The photographs' standardization at a resolution of 1080 x 1920 pixels makes them uniform and easy to use for a range of applications. Each of the three sets of leaves in this large dataset indicates a distinct aspect of lychee health and vitality. 1. Diseased Lychee Leaves (Class I): Total Count: 3,539 images Subcategories: This extensive class further subdivides into six subcategories, each representing a specific lychee leaf disease: -Entomosporium Leaf Spot on Woody Ornamentals - Leaf Blight Lychee Leaf Diseases - Lychee Algal Spot in Non-Direct Sunlight - Lychee Anthracnose on Cloudy Day - Lychee Leaf Mites in Direct Sunlight - Lychee Mayetiola After Raining 2. Dry Lychee Leaves (Class II): Total Count: 476 images Description: This category comprises images of lychee leaves in a desiccated state, reflecting the natural progression of foliage as it transitions from vigor to senescence. The dry lychee leaves may showcase characteristic features such as wilting, discoloration, and a withered appearance. 3. Healthy Lychee Leaves (Class III): Total Count: 114 images Description: These pristine specimens exemplify the ideal state of lychee leaves, exhibiting vibrant green hues, uniform textures, and the absence of any visible blemishes. They serve as a benchmark for assessing the well-being of lychee foliage.



Daffodil International University


Agricultural Engineering, Data Science