The Udila Tea Leaves Dataset: Enabling Deep Learning Classification for Premium Tea Production

This dataset contains a curated collection of tea leaf images sourced from Udila Tea Garden in Chittagong, Bangladesh, specifically designed to classify leaves based on their suitability for high-quality tea production. The dataset is categorized into two classes: "Tea-Making Leaf" and "Not Tea-Making Leaf" with a total of 747 samples. The "Good for Tea-Making" class includes fresh, vibrant leaves identified as ideal for producing flavorful, premium tea. In contrast, the "Not Good for Tea-Making" class contains leaves that are either aged, unhealthy, or otherwise unsuitable for quality tea production. Data Collection: Images were captured under controlled natural lighting to accurately reflect the visual characteristics of each leaf. "Good" leaves are typically new, fresh, and exhibit a desirable green color and healthy structure. "Not Good" leaves may show signs of aging, poor health, or other qualities that would compromise the tea flavor and quality. Applications: This dataset can be used for machine learning models aimed at quality assessment in tea production, enabling automated sorting and decision-making for optimal tea leaf selection in agricultural and commercial tea production processes.