Carotene content analysis using spectrophotometer and artificial neural network: article / Arfah Saman

This paper presents a study to investigate the carotene content of palm oil fruit using spectrophotometer and Artificial Neural Network (ANN). The study confined only three different ripeness of palm oil fruit on Tenera species that are known as ripe, over-ripe and under-ripe. 10 samples of each ripeness of fruit are analyzed and scanned to capture image spectrum through spectrophotometer. Then, in turn that image will be used in the analysis of backpropagation neural network in determining the highest carotene from ripeness of palm oil fruit. The result shows that ripe is the best ripeness of palm oil which contain highest carotene through errors analysis from actual and target outputs out of 30 samples of difference ripeness each. From that analysis, it can be conclude that smaller error represent more quality carotene content of palm oil.