Development and Study of a Garlic Bulb Sorting Machine
蒜球 ； 分級 ； 影像處理 ； Garlic bulbs ； Grading ； Image processing
|Volume or Term/Year and Month of Publication||
|Academic Degree Category||
目前坊間蒜球分級多使用人工以目視方式來進行分級作業，此分級方式耗力費時且不夠客觀，故本研究旨在研製一套蒜球分級系統，期能取代人工分級，快速且準確地完成分級作業，降低收穫後處理成本、提高蒜球收購價格與增加蒜農收益，以利後續農事單位收購與儲藏。 本研究先行以影像處理等方式來進行蒜球基礎物性調查，結果顯示蒜球外形與重量具有高度相關性，故設計一套以外形為選別機制之分級機，階級式蒜球分級機所採用之分級原理，為利用改變階級毛刷間相對應位置，即分級間隙來達到分級的目的。而蒜球分級系統係由自動供料機構、去皮脫膜機構與階級式蒜球分級機等所組成。 利用不同蒜球莖長與分級轉速試驗得知，分級準確率分佈在79.3~90.1%，而以蒜球莖長3公分，分級轉速86rpm時，分級準確率最高為90.1%，而在中華民國國家標準(CNS)所給予容許5~10%誤差範圍時，其平均分級準確率可達98.7%，分級作業量每小時約576公斤，較單排側放式與雙排下放式重量分級機分別高 2.2與1.4倍；且階級式蒜球分級機售價與分級作業成本，亦低於單排與雙排式重量分級機，故使用階級式蒜球分級機可節省相當之人力與成本。
At present, the traditional grading method of the post-harvest garlic bulbs used manpower with eye vision that was time consumption and subjective. The purpose of this research aimed on the development of a garlic bulbs sorting system to replace the traditional grading method. The goal of the system was to achieve the fast and accurate grading operation so as to reduce the post-harvest cost, increase the garlic bulbs purchasing price, raise the farmers income, and help the following wholesaler purchase and storage. This research first applied the image processing technique to investigate the basic physical characteristics of the garlic bulbs. The results studied showed that the shape size and weight of the garlic bulbs held with high correlation. Therefore, a shape sorting mechanism was designed which utilized the change of different spacing sections of two hairbrush shafts by adjusting their relative positions as operation under the same rotation speed and direction. The system developed for the garlic bulbs grading was combined together with the parts of the automatic supplying unit, the peel and molt removing equipment, and the shape sorting mechanism. Based on the testing results of this newly developed sorting system with various stem length of the garlic bulbs and various rotation speed of the hairbrush shafts, the accuracy of grading obtained was between 79.3% and 90.1%. The condition of the case with the maximum accuracy (90.1%) was with the stem length 3 cm and rotation speed 86 rpm. Under the allowable tolerance errors 5~10% described in the Chinese National Standards(CNS), the average grading accuracy then obtained of the whole testing cases could be raised up to 98.7%. The processing capacity of the system was reached up to about 576 kg per hour which was 2.2 and 1.4 times, respectively as compared to the system obtained with the single-rail and double-rail weight sorting mechanism. Moreover, the newly developed shape sorting system also held the lower construction expenditure and operation cost than the single-rail and double-rail weight sorting machine. It could save tremendous labor power and processing cost as required in the grading operation of the garlic bulbs.
生物農學 > 生物環境與多樣性
工程學 > 電機工程