Classification of Oil Palm Fresh Fruit Bunches Based on Their Maturity Using Thermal Imaging Technique

The maturity of oil palm Fresh Fruit Bunches (FFB) is considered to be a significant factor that affects the profitability and salability FFB. Typical methods grading FFB consist physical fresh fruit, which time-consuming expensive, results are prone human error. Therefore, this research attempts formulate thermal imaging method indicate precise fruits. A total 297 FFBs were collected. samples divided into three groups: under-ripe, ripe, over-ripe. Afterward, all scanned using camera calculate real temperature each sample. In order normalize measurement, difference between average bunch ambient (∆Temp) was as main parameter. indicated mean ∆Temp decreased consistently from under-ripe ANOVA test demonstrated observed significance value less than 0.05 in terms ∆Temp, so there statistically means categories. It can concluded reliable index classify palm. classification analysis conducted its application an Linear Discriminant Analysis (LDA), Mahalanobis (MDA), Artificial Neural Network (ANN), Kernel Nearest Neighbor (KNN). highest degrees overall accuracy (99.1% 92.5%) obtained through ANN method. This study concludes images used classification.