Nature's Diabetes Helpers: Insulin Leaves Repository

This database comprises a total of 429 images depicting both healthy (265 images) and diseased (164 images) leaves of the insulin plant (Scientific name: Chamaecostus cuspidatus). All the images included in this database are captured by the creators of this database from the real fields located in the Ranaghat district of West Bengal, India. The images in this database consist of intricate backgrounds that include elements like weeds, soil, and portions of other plants. The primary objective behind creating this database is to offer a comprehensive assortment of labeled foliage images of insulin plants. These images are intended for training deep neural networks to autonomously classify them into either diseased or healthy categories. The incorporation of foliage images with complex backgrounds and captured under varied illumination condition in this database enhances its capability to effectively train deep neural networks, enabling precise real-world identification for the specified task. The images included in this database were captured using mobile phones, specifically the Redmi Note 9 Pro Max with a camera array featuring 64MP+8MP+2MP+2MP and Redmi Note 9 Pro Max with a camera array featuring 64MP+8MP+5MP+2MP, utilizing Artificial Intelligence (AI) technology. As this database contains the diseased and healthy leaves of insulin plants which serves as natural remedy for diabetes, so we have termed this database as “Nature's Diabetes Helpers: Insulin Leaves Repository.”