Implementation K-Nearest Neighbor (KNN) Algorithm for Identifying Pattern Differences Between Betel Leaves and Pepper Leaves

This research aims to identify the pattern differences between betel leaf (Piper betle) and black pepper leaf (Piper nigrum) using the K-Nearest Neighbor (KNN) algorithm. The method is applied to classify the two types of leaves based on image features such as texture, shape, and color. A total of 200 leaf images were used, divided into 70% training data and 30% testing data. Feature extraction was conducted to obtain the most relevant characteristics from each image. The classification process was performed with various K values, and the highest accuracy was achieved when K = 3. The results showed an accuracy of 91.5%, with a precision of 90.8%, recall of 92.3%, and F1-score of 91.5%. These findings indicate that the KNN algorithm is effective in distinguishing betel and pepper leaves using digital image processing. Texture and color features contributed the most to the classification performance. This study shows the potential of KNN-based leaf pattern recognition for practical applications in agriculture, herbal identification, and plant classification systems.