Jingwei Wang , Xiaopeng Bai , Daochun Xu , Wenbin Li , Siyuan Tong , Jiaming Zhang
{"title":"Online sorting of surface defective walnuts based on deep learning","authors":"Jingwei Wang , Xiaopeng Bai , Daochun Xu , Wenbin Li , Siyuan Tong , Jiaming Zhang","doi":"10.1016/j.jfoodeng.2024.112133","DOIUrl":null,"url":null,"abstract":"<div><p>To address concerns regarding walnut shell damage and inadequate sorting precision during the mechanized sorting of walnuts, a walnut automatic sorting machine was designed based on deep learning and experimental research. Initially, the rationality of the design was verified through experiment. Then, three deep learning semantic segmentation algorithms, namely PSPnet, U-net, and Deeplabv3+, were selected to train walnut detection models. Results indicated that the U-net algorithm proved to be the most effective, achieving a Mean Intersection over Union of 96.71% and a Mean Pixel Accuracy value of 98.52%. Finally, performance tests were conducted on the prototype machine, yielding results with an average sorting efficiency of 51.70 kg/h, an average loss rate of 6.50%, and an average accuracy of sorting walnuts of 92.98%. The findings can provide insights for future structural improvements and operational parameter optimization of walnut automatic sorting machines.</p></div>","PeriodicalId":359,"journal":{"name":"Journal of Food Engineering","volume":null,"pages":null},"PeriodicalIF":5.3000,"publicationDate":"2024-05-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Food Engineering","FirstCategoryId":"97","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0260877424001997","RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CHEMICAL","Score":null,"Total":0}
引用次数: 0
Abstract
To address concerns regarding walnut shell damage and inadequate sorting precision during the mechanized sorting of walnuts, a walnut automatic sorting machine was designed based on deep learning and experimental research. Initially, the rationality of the design was verified through experiment. Then, three deep learning semantic segmentation algorithms, namely PSPnet, U-net, and Deeplabv3+, were selected to train walnut detection models. Results indicated that the U-net algorithm proved to be the most effective, achieving a Mean Intersection over Union of 96.71% and a Mean Pixel Accuracy value of 98.52%. Finally, performance tests were conducted on the prototype machine, yielding results with an average sorting efficiency of 51.70 kg/h, an average loss rate of 6.50%, and an average accuracy of sorting walnuts of 92.98%. The findings can provide insights for future structural improvements and operational parameter optimization of walnut automatic sorting machines.
期刊介绍:
The journal publishes original research and review papers on any subject at the interface between food and engineering, particularly those of relevance to industry, including:
Engineering properties of foods, food physics and physical chemistry; processing, measurement, control, packaging, storage and distribution; engineering aspects of the design and production of novel foods and of food service and catering; design and operation of food processes, plant and equipment; economics of food engineering, including the economics of alternative processes.
Accounts of food engineering achievements are of particular value.