Date Fruits Grading and Sorting Classification Algoritham Using Colors and Shape Features

Abstract

The kingdom of Saudi Arabia produces almost 400 varieties, is the largest producer of dates in the worlds in dates fruit industries. The date's industries have facing problems in date grading and sorting during harvesting and agriculture products industries. Since human experts have some limitations such as tedious and timing delay consuming the processing delay as a result of date products costly. The computer vision algorithms are used to reduce the powers of mankind by automatically detect and classify an image. The main aim of the computer vision-based algorithms for data fruits grading and sorting is to make the procedure fully automated and reduce human interference. The computer vision algorithm uses RGB pictures, which automatically abstracts the aforesaid exterior date class arrangements. Based on the extracted algorithms, the proposed vision-based machine automatically classifies dates into three different quality based categories (class A, B, and C) defined by the user. The paper studied the accuracy of the computer vision-based algorithm based on pre-selected date fruits samples from different field in Al-Qassim, Sudi Arabia. The experimental grades classification illustrated that the proposed algorithm is able to sort the date's fruits up to 99% precisely.

A Deep-Learning Model for Real-Time Red Palm Weevil Detection and Localization

“…Several imaging techniques have been proposed that rely on template matching to identify the pest [ 12 ]. An IoT-based smart palm-weevil monitoring system was developed based on using a web/mobile interface to detect the red palm weevil via sensors [ 1 ]. By applying 10 state-of- the-art data mining algorithms for classifications, tremendous work was done in 2021.…”

“…The limits of the proposals are also adjusted with respect to linking boxes by regression. Second, for each region proposal, the features within the region are first aggregated into a feature map of constant size (i.e., RoI pooling [ 1 ]). Using the pooling feature, a DNN classifier then computes the probabilities of object classes and simultaneously regresses the detection limits for each object class.…”

“…In the early part of the 20th century, its presence was recognized as a medical condition in Southeast Asia. Furthermore, its presence has caused damage in the western and eastern parts of Asia as well as in northern Africa and Europe [ 1 , 2 ]. The RPW had spread by the end of the 20th century and was discovered in the western parts of Northern America by the end of the decade.…”

Date Fruit Sorting Based on Deep Learning and Discriminant Correlation Analysis

“…To deal with the problem of limited number of labeled images, authors in [28] suggested using an evolutionary algorithm along with an AdaBoost training procedure to check the quality of date fruit. In [29] , a hand-designed basic color and shape features were utilized to classify dates into three classes of quality namely class A, B and C. In another research [30], Mozafati dates were classified, based on length and freshness parameters, into different categories, from very poor to the excellent quality.…”