Non-Destructive Classification of Fruits Based on Vis-nir Spectroscopy and Principal Component Analysis

Abstract

Fruits are one of the sources of nutrition needed for health. Fruit quality is generally assessed by physical and chemical properties. Measurement of fruit internal quality is usually done by destructive techniques. Ultraviolet, visible and near-infrared (UV-Vis-NIR) spec-troscopy is a non-destructive technique to measure fruit quality. This technique can rapidly measure the fruit quality, the measured fruit still remains intact, and can be marketed. Besides, UV-Vis-NIR spectrosco-py can also be used to classify fruits. The study aimed to classify var-ious types of fruits using UV-Vis-NIR spectroscopy with wavelengths of 300-1041 nm and Principal Component Analysis (PCA). First de-rivative savitzky-golay with 9 smoothing points (dg1) and multiplica-tive scatter correction (MSC) were applied to correct the spectra. The results showed that the use of uv-vis-nir spectroscopy and PCA com-bined with spectra pre-treatment of the MSC method were able to clas-sify various types of fruits with 100% success rate in all fruit samples including sapodilla, ridge gourd, mango, guava, apple and zucchini.

YOLOv4-Driven Appearance Grading Filing Mechanism: Toward a High-Accuracy Tomato Grading Model through a Deep-Learning Framework

“…Naik and Patel reviewed the literature on fruit classification and grading [27]. They explored the classification and grading capabilities of popular machine learning methods such as artificial neural networks [28], biogeography-based optimization [29], support vector machines [30], and principal component analysis [31] under different fruits. Their results confirmed that machine learning could achieve fruit detection, tracking, and estimation.…”

PENERAPAN ALGORITMA DISKRIMINASI MENGGUNAKAN METODE PRINCIPAL COMPONENT ANALYSIS (PCA) DAN Vis-SWNIR SPECTROSCOPY PADA BUAH CABAI RAWIT DOMBA BERBAGAI TINGKAT KEMATANGAN

“…PCA berfokus pada berbagai variasi dari spektrum dan mengubahnya menjadi varibel-variabel baru yang tidak saling berkorelasi. PCA telah banyak digunakan untuk mengelompokkan berbagai produk pertanian [7]- [11]. Metode PCA dapat diaplikasikan salah satu tujuannya yaitu untuk grading komoditas hortikultura.…”

“…Hal ini menjelaskan bahwa sebanyak 90% data keragaman dapat dijabarkan di PC1 dan 7% data keragaman dapat dijelaskan di PC2. Sejalan dengan penelitian lainnya yang menyebutkan bahwa dengan metode PCA dan spectroscopy didapatkan keakuratan 100% pada berbagai produk khususnya dalam bidang pertanian [7], [9], [16].…”