کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
5762626 | 1624959 | 2018 | 13 صفحه PDF | دانلود رایگان |
عنوان انگلیسی مقاله ISI
Comparative study of transform-based image texture analysis for the evaluation of banana quality using an optical backscattering system
ترجمه فارسی عنوان
بررسی مقایسه ای از تجزیه و تحلیل بافت تصویر مبتنی بر تبدیل برای ارزیابی کیفیت موز با استفاده از سیستم برگشت ناپذیر نوری
دانلود مقاله + سفارش ترجمه
دانلود مقاله ISI انگلیسی
رایگان برای ایرانیان
کلمات کلیدی
تبدیل موجک، تبدیل گابور، تبدیل تامورا، شبکه های عصبی مصنوعی، ماشین بردار پشتیبانی،
موضوعات مرتبط
علوم زیستی و بیوفناوری
علوم کشاورزی و بیولوژیک
علوم زراعت و اصلاح نباتات
چکیده انگلیسی
In this study, the application of a backscattering imaging system with different approaches of transform-based image texture analysis for the evaluation of banana quality at different ripening stages was investigated with Wavelet, Gabor and Tamura transforms. The attenuated images of the fruits were acquired using Laser Light Backscattering Imaging (LLBI) with laser diodes emitting light at three wavelengths viz 532, 660, and 830 nm. The elasticity, chlorophyll index and soluble solids content (SSC) of each sample were measured as reference parameters by using a texture analyser, a Delta Absorbance (DA) meter, and a refractometer, respectively. The performance of the extracted features from the selected transform-based image texture analysis for analysing the quality parameters of the fruit was evaluated by means of an artificial neural network (ANN) and a support vector machine (SVM). The results indicated that there were significant changes of elasticity, chlorophyll index and SSC as the ripening stages increased. Prediction model analysis showed that the Wavelet transform exhibited the most reliable results for all of the reference parameters followed by Tamura and the Gabor transform. The results also revealed that analysis using an ANN approach recorded better performance than SVM as reflected by higher coefficient of determination (R2) values. Thus, this study indicated that an LLBI system with transform-based image texture analysis coupled with computational intelligence techniques can be used for the evaluation of the quality of bananas.
ناشر
Database: Elsevier - ScienceDirect (ساینس دایرکت)
Journal: Postharvest Biology and Technology - Volume 135, January 2018, Pages 38-50
Journal: Postharvest Biology and Technology - Volume 135, January 2018, Pages 38-50
نویسندگان
Norhashila Hashim, Segun Emmanuel Adebayo, Khalina Abdan, Marsyita Hanafi,