The identification of millets is important for ensuring food security and sustainable agriculture. This work proposes a novel approach that combines computer vision techniques with ML algorithms to recognize different types of millets based on their visual features. These features are used to train a ML framework that can classify millets with high accuracy. The proposed system used 3 samples of millets which include jowar, bajra and raagi. Based on their shape and texture features the model predicts the millets using various machine learning algorithm. The accuracy of the proposed work is 92% using random forest classifier.