Oil palm fresh fruit bunch ripeness classification using back propagation and learning vector quantization

3rd Annual Applied Science and Engineering Conference

工业技术;自然科学

Fahmi, F.^1 ; Palti, H.^1 ; Emerson, S.^1 ; Suherman, S.^1

Electrical Engineering Department, Universitas Sumatera Utara, Medan, Indonesia^1

关键词: Classification results;    Fresh fruits;    Fruit samples;    Learning Vector Quantization;    Learning vectors;    Research interests;    Simple analysis;    Unripe fruits;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/434/1/012066/pdf
DOI  :  10.1088/1757-899X/434/1/012066

来源: IOP

【 摘 要 】

Fresh fruit bunch analysis has been research interest for many years. Various techniques have been proposed. However, complex techniques may exert problem in implementation, This article report the fresh fruit bunch ripeness identification by using back propagation and learning vector quantification to identify whether the fruits ripen or not. Simple analysis methods are used so that application such as drone based identification can be easily implemented. The fruit sample contains fresh ripe fruit bunch (RFB) and fresh unripe fruit bunch (UFB). By using 20 RFBs and 20 UFBs, the classification results at least 95% precision, 98% accuracy, sensitivity 1, and specificity 0.95.