DETECTION OF BLACK MOLD INFECTED FIGS BY USING TRANSMITTANCE SPECTROSCOPY


Durmus E., Bilgi A. S. , Ortac G., Kalkan H., Tasdemir K.

7th Workshop on Hyperspectral Image and Signal Processing - Evolution in Remote Sensing (WHISPERS), Tokyo, Japan, 2 - 05 June 2015 identifier identifier

Abstract

Environmental conditions (humidity, temperature, wind, etc.) and inappropriate processing and storage conditions effects the quality of agricultural products. Figs, like other agricultural products, are mostly affected by molds during drying and processing and the damage given by molds are hardly detected by visual controls. Aspergillus niger is a type of mold that penetrates into figs and turns the color into black when grows inside figs and that black structures cannot be usually observed from outside. These figs are usually detected by nailing method, which depends on penetrating a needle into the figs and taking some specimen to examine visually. However, this procedure is labor expensive and includes the risk of transmitting the molds to the sound figs. In this study, we propose a non-destructive and fast method to detect the black-mold contaminated figs by using transmittance spectroscopy. Figs are screened at 3648 spectral bands between 200 nm-1100 nm and classified by using the most relevant spectral bands. By using the most discriminative 5 spectral bands, 100% classification accuracy has been achieved.