Perancangan Aplikasi Indentifikasi Kematangan Jambu Madu Dengan Metode Ekstraksi Ciri Statistik

Abstract

Guava Honey is a fruit that can be consumed and has a wide market share. The similarity in texture of honey guava skin color between ripe and immature results in difficulty in identifying ripe honey guava in terms of fruit skin texture characteristics and subjective human assessments of the maturity level of Honey guava fruit causing the assessment of the maturity level of Guava Honey to differ from one with others. From these problems, so that research was carried out to identify the maturity of Guava Honey based on the texture of the fruit skin color. The purpose of this study was to apply the statistical feature extraction method with characteristic parameters, namely Mean (μ), Variance (σ2), Skewness (α3), Kurtosis (α4) and Entropy (H) as a method to recognize the ripeness of honey guava in terms of fruit skin texture. and to find out the accuracy value after the system is tested. The subject of the research was to build an application to identify the ripeness of honey guava to statistically detect the ripe fruit of the ripe honey guava in terms of the texture of the fruit skin. Based on this research, the authors designed an Android-based application of guava honey maturity identification using Eclipse Juno and Sqlite. To simplify the identification process in the application, the writer applies the Statistical Feature Extraction Method and the expected results in this study are an application that applies the Statistical Feature Extraction Method.