Open AccessIdentification of male and female nutmeg seeds based on the shape and texture features of leaf images using the learning vector quantization (LVQ)
Author(s) -
examiner Yudi Satria,
Fitri Arnia,
Kahlil Muchtar
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1807/1/012001
Subject(s) - learning vector quantization , nutmeg , mathematics , artificial intelligence , pattern recognition (psychology) , machine learning , vector quantization , computer science , biology , food science
One of the stages in the process of planting nutmeg seedlings on agricultural land is the identification of the sex of male and female nutmeg seedlings. The process of identifying nutmeg seedlings still uses manual methods, while the method is often inaccurate and inefficient. There fore, this research discusses the methods and results of identification of male and female nutmeg seedling leaves based on the combined feature extraction and leaf texture using learning vector quantization (LVQ). The form feature extraction method used consists of slimness, roundness, narrow factor, perimeter ratio and diameter, perimeter ratio at length while for texture is contrast, correlation, energy, and homogeneity. The results of the combined extraction of shape and texture features were applied as input vectors when classification using the LVQ method. Whereas LVQ aims at classification based on extraction results, by initializing LVQ parameters, namely learning rate and epoch. The change in the learning rate value on LVQ is very influential to get the percentage of the correctness of the data, the learning rate value must be between 0 to 1. The accuracy of LVQ predictions obtained in this study is 92.30%, with the best learning rate values from 0.001 to 0.09 and epoch 50.