In this paper, an image processing technique for estimating the
volume and mass of four citrus fruits was presented and discussed.
The method is quite general and may be readily applied for volume
and mass computation of other axi-symmetric agricultural products
such as carrot, cucumber, onion, melon, kiwifruit, pomegranate,
and pear. Based on the results presented in this paper, we can
argue the proposed method is rotationally invariant and does not
require fruit alignment on the conveyor.
The present work may be extended in several ways. For example,
the surface area of the citrus fruits can also be computed in a
similar way (Khojastehnazhand et al., 2009). The background segmentation
method adapted here is not based on threshold values,
and can readily be adapted for other fruits.
The method developed in this study provides an alternative to
the traditional methods for measurement of volume, surface area,
and mass of axi-symmetric agricultural products. The estimated
volume, surface area, and mass of fruits, and the already available
RGB color information may be used for online sorting of various citrus
fruits. However, RGB system adapted in this paper could be
sensitive to lighting or other conditions. Therefore, we have used
HSI color space instead to develop a sorting device for grading citrus
(Khojastehnazhand et al., 2008). In the HSI system, hue value is
comparatively stable and the color of citrus can be determined by
calculating the average Hue value for the fruit. Works in this direction
is still in progress.
In this paper, an image processing technique for estimating thevolume and mass of four citrus fruits was presented and discussed.The method is quite general and may be readily applied for volumeand mass computation of other axi-symmetric agricultural productssuch as carrot, cucumber, onion, melon, kiwifruit, pomegranate,and pear. Based on the results presented in this paper, we canargue the proposed method is rotationally invariant and does notrequire fruit alignment on the conveyor.The present work may be extended in several ways. For example,the surface area of the citrus fruits can also be computed in asimilar way (Khojastehnazhand et al., 2009). The background segmentationmethod adapted here is not based on threshold values,and can readily be adapted for other fruits.The method developed in this study provides an alternative tothe traditional methods for measurement of volume, surface area,and mass of axi-symmetric agricultural products. The estimatedvolume, surface area, and mass of fruits, and the already availableRGB color information may be used for online sorting of various citrusfruits. However, RGB system adapted in this paper could besensitive to lighting or other conditions. Therefore, we have usedHSI color space instead to develop a sorting device for grading citrus(Khojastehnazhand et al., 2008). In the HSI system, hue value iscomparatively stable and the color of citrus can be determined bycalculating the average Hue value for the fruit. Works in this directionis still in progress.
การแปล กรุณารอสักครู่..
