AI-Powered Plant Pathology: Unveiling Tea Leaf Diseases with a Comprehensive Dataset
Description
1.The Significance of tea leaf in Horticulture and Disease Challenges Tea leaves are highly prized in the horticultural sector for their aesthetic worth as well as their economic relevance. However, a number of illnesses that can negatively affect the health and production of these plants frequently provide a barrier to their cultivation. To reduce these hazards, leaf disease identification must be done accurately and early. In response to this requirement, the Tea Leaf Disease Dataset was created to help researchers, horticulturists, and machine learning specialists recognize and categorize illnesses that impact tea leaves. High-resolution photos of tea leaves, showing both healthy and sick specimens, are included in this collection. The dataset contains notable illnesses including anthracnose and gray blight. Precise model training and validation are made possible by the thorough annotation of every image with information on the nature and severity of the illness. Metadata that offers crucial contextual information about the locations and weather at the time of image collection is added to the dataset, further enriching it. This contextual information improves the accuracy of prediction models and is essential for comprehending the factors driving the development of illness. 2. The Role of Deep Learning and Computer Vision The tea leaf Disease Dataset has a plethora of opportunities for classification and detection tasks that can be addressed by modern deep learning and computer vision techniques. 3. A Comprehensive Dataset for Deep Learning in tea leaf Disease Classification This dataset was created specially to aid in the creation of sophisticated deep learning models. To ensure accuracy and relevance, the dataset's classifications were developed in conjunction with subject matter specialists from a leading agricultural institute. 4. Data Collection and Augmentation At the Allynugger Duncan Tea Garden site in Moulvibazar, Sylhet, Bangladesh, a total of 3900 photos were captured. These images showed examples of healthy leaves, anthracnose, and gray blight. 1900 augmented pictures were produced from the initial photo collection in order to enhance the dataset and increase the number of data points that were accessible. Several methods, including flipping, rotating, shearing, brightness alteration, and magnification, were used to create these enhanced samples.