Comparison of pyrus fruit classification using K-Nearest Neighbor (KNN) and Adaptive Neuro Fuzzy Inference System (ANFIS) algorithms

Pears or Pyrus fruits are popular among the public because of their high nutritional content, delicious taste and low calories. Fruit classification can be done visually, but manual classification requires consistent techniques and is often constrained by human aspects. Digital image processing is applied to solve the above problems. The pears identified in this study came from three types, namely the Abate Pear, the Monster Pear and the William Pear. Pre-processing is done by converting the RGB image to L*a*b, then segmenting it using the K-Means Clustering algorithm. The segmented image is extracted into 7 features, namely 6 color features (Red, Green, Blue, Hue, Saturation, Value) and 1 size feature (Area). Classification is carried out using the K-Nearest Neighbor (KNN) algorithm and the Adaptive Neuro Fuzzy Inference System (ANFIS). The results showed that the KNN algorithm had a better performance in classifying the types of pears.