MangoDHDS: Dataset for Mango Fruit Disease Detection Comprising 1.5K Images

Mango fruit diseases have a huge impact on agriculture, diminishing yields and quality, resulting in substantial marketing and economic losses. Mango, a cultivated fruit in India, requires efficient practices and increased yields to benefit the agricultural and commercial sectors. Mango is a national fruit that originated in South East Asia's Indo-Burma region. The Mango Diseased Healthy Dataset, named as MangoDHDS, is created by combining diseased and healthy mango fruit samples from various sources. This is labelled image dataset intended for training machine learning models to detect diseases in mango fruits. This dataset is comprised a meticulously organized collection of 1,500 images of mango fruits, classified into five categories: Anthracnose, Bacterial Canker, Scab, Stem End Rot, Healthy. Dataset containing images of 256*256 in JPG format. Diseased and healthy mango images are captured from various sources, such as research centers, ripening chambers, agricultural farms, and Jain Farm Fresh Foods Limited, Jalgaon, Dr. Balasaheb Sawant Konkan Krishi Vidyapeeth, Dapoli, and Kanhere Agro Products, Dapoli (India). Mangoes are a seasonal fruit; therefore, all images were taken between April and July from 2021 to 2023. All images were captured using smart phone camera with white background. MangoDHDS includes both preharvest and postharvest diseases. The collection was curated by manually choosing clear and verifying proper categorization for each disease category under the supervision of an expert. Each class has an equal number of images; therefore, the dataset is well-balanced and suited for supervised machine learning applications. The dataset is useful for developing automated systems for tasks like disease detection through image analysis, classification using real world fruit images, fruit disease diagnosis applications, smart farming, recommendations according to diseases, calculating disease severity. Dataset is made available to support model evaluation, and further research in the fields of image processing, machine learning, deep learning, agriculture and computer vision.We strongly recommend that interested researchers use this dataset for image processing and machine learning applications to recognize mango fruit diseases.