This paper focuses on two varieties of date fruits in the tamr stage. Bomaan and Lolo are
among the most common date fruits in the Gulf area. Ten date fruits of each variety were
picked from the El-Sad factory randomly. The image of each fruit was captured using the
Mavica digital camera under fluorescent illumination. Matlab of Mathworks was used to find
out the R, G, and B intensity of each pixel of each fruit’s image and then to calculate the mean
value for each fruit. The fructose and glucose content of each fruit were determined using the
HPLC technique. Since sucrose decomposes into mono sugars under storage conditions, it
was neglected in this study. Furthermore, HPLC results showed that sucrose content is less
than 0.2% in the whole sample. SPSS regression tools were used to relate each fruit fructose
and glucose content to its associate R and G intensity mean value. As a result, prediction
equations of fructose and glucose content according to R or G intensity were developed for
each variety. For validation, additional samples of 43 fruits of the two varieties were
collected from two different sources. 31 of them were gathered from the El-Sad factory and
the rest were collected from the local market were nobody could trace their producer. The
fruits in the additional samples were handled in the same way the first sample’s fruits were
handled. The one exception was that the values of color intensity were applied in the
associated prediction equation and the predicted values of sugar content were compared to the
resulting values from the HPLC system. SPSS was used to find out the prediction error for
each variety.
This paper focuses on two varieties of date fruits in the tamr stage. Bomaan and Lolo areamong the most common date fruits in the Gulf area. Ten date fruits of each variety werepicked from the El-Sad factory randomly. The image of each fruit was captured using theMavica digital camera under fluorescent illumination. Matlab of Mathworks was used to findout the R, G, and B intensity of each pixel of each fruit’s image and then to calculate the meanvalue for each fruit. The fructose and glucose content of each fruit were determined using theHPLC technique. Since sucrose decomposes into mono sugars under storage conditions, itwas neglected in this study. Furthermore, HPLC results showed that sucrose content is lessthan 0.2% in the whole sample. SPSS regression tools were used to relate each fruit fructoseand glucose content to its associate R and G intensity mean value. As a result, predictionequations of fructose and glucose content according to R or G intensity were developed foreach variety. For validation, additional samples of 43 fruits of the two varieties werecollected from two different sources. 31 of them were gathered from the El-Sad factory andthe rest were collected from the local market were nobody could trace their producer. Thefruits in the additional samples were handled in the same way the first sample’s fruits werehandled. The one exception was that the values of color intensity were applied in theassociated prediction equation and the predicted values of sugar content were compared to theresulting values from the HPLC system. SPSS was used to find out the prediction error foreach variety.
การแปล กรุณารอสักครู่..
