A Decision-Support for Rubber Tree Clone Diagnosis using Fuzzy Logic

Initial classification of the rubber tree clone series at the early stage is a crucial decision as there are a lot of clones that need to be chosen for plantation purposes. This research focuses on developing a decision-support for rubber tree clone diagnosis based on optical properties using Fuzzy logic. This work selected three RRIM clones: RRIM 2007, RRIM 2014, and RRIM 3001, based on the most excellent latex general remarks with good latex yielding and suitable for timber products. The total samples selected consist of 480 samples and 50 samples from each clone for the development of the system and validation, respectively. The prior researcher extracted the data using the developed optical sensor with RGB LED and simple statistical analyses to find the correlation and relationship between RGB spectrum and clone properties. The preliminary statistical data, such as information on the descriptive; box plots of the RGB spectrum, are needed to use as the membership inputs and the clone type as the outputs in the Fuzzy logic system. The decision system is a success, with findings showing that the model can adequately validate RRIM 2014 with 98% accuracy, compared to 76% and 56% for RRIM 3001 and RRIM 2007, respectively.