Lentil Classification Dataset: Polished and Unpolished

Realizing the increasing demand for quality data in food quality testing, agricultural research, and retail research, the dataset was created to record fine visual variations between polished and unpolished lentils. It is centered on five major varieties of lentils commonly eaten in India: Chana, Masoor, Moong, Toor, and Urad. The dataset includes a total of 1097 high-resolution images (3005 × 3005 pixels) of these lentils, which were captured specifically to differentiate between their polished and unpolished states. Each picture is systematically titled: for unpolished specimens as lentilname_UP_001 and for polished ones as lentilname_P_001, where "lentilname" indicates the lentil variety (Chana, Masoor, Moong, Toor, Urad), "UP" indicates the polish status ("UP" for Unpolished and "P" for Polished), and the number indicates the sequence of the pictures. All images were taken from nearby retail shops under standardized light conditions and identical background settings to ensure consistency and guarantee high quality of inspection.