基于主动热红外成像的马铃薯与杂质分类方法  

Classification of potato and impurities based on active thermal infrared imaging

作　　者：孙卫孝 刘发英[2] 杨振宇[1,3] 韩梦杰 李学强 魏忠彩 Sun Weixiao;Liu Faying;Yang Zhenyu;Han Mengjie;Li Xueqiang;Wei Zhongcai(School of Mechanical Engineering,Shandong University of Technology,Zibo,255000,China;School of Electrical and Electronic Engineering,Shandong University of Technology,Zibo,255000,China;Shandong Provincial Intelligent Engineering and Technology Research Center of Potato Production Equipment,Dezhou,253600,China;School of Agricultural Engineering and Food Science,Shandong University of Technology,Zibo,255000,China)

机构地区：[1]山东理工大学机械工程学院,山东淄博255000 [2]山东理工大学电气与电子工程学院,山东淄博255000 [3]山东省马铃薯生产装备智能化工程技术研究中心,山东德州253600 [4]山东理工大学农业工程与食品科学学院,山东淄博255000

出　　处：《中国农机化学报》2024年第2期143-150,共8页Journal of Chinese Agricultural Mechanization

基　　金：山东省科技型中小企业创新能力提升工程项目(2021TSGC1420);山东省农业重大应用技术创新项目(SD2019NJ010);国家自然科学基金项目(52105266);中国博士后科学基金面上资助项目(2021M701801)。

摘　　要：为解决入库马铃薯中石块、土块等杂质剔除的问题,提出一种基于主动热红外成像的马铃薯与杂质分类方法。建立基于集总参数法的马铃薯和杂质的传热模型,获得影响受热效果与导致温度差异的因素为材料热物性、风速、风向角度、风温;利用有限元法分析马铃薯和杂质在外部热风激励下的热像图,马铃薯表面温度显著高于石块与土块,并对风速、风向角度、风温分别进行四水平单因素仿真试验,得出马铃薯与杂质的表面温度差异大小在一定时间内与风速、风向角度和风温呈非线性正相关。在试验平台上对马铃薯和杂质进行分类试验,结果表明:最佳图像识别处理条件为风速4 m/s、风温为40℃、风向角度为90°,识别成功率为97%。为马铃薯收获—仓储过程中除杂提供技术参考。Aiming at the problem of removing impurities such as rocks and clods from stored potatoes,a classification method of potato and impurities based on active thermal infrared imaging was proposed.Firstly,the heat transfer model of potato and impurity was established based on lumped parameter method,and the factors affecting the heating effect and leading to temperature difference were obtained as material thermal properties,wind speed,wind direction angle and wind temperature.Then,using the finite element method(FEM)to analyze the thermal image of potatoes and impurities under the hot air excitation,the potato surface temperature was significantly higher than that of rocks and clods,and four level single factor simulation experiment was carried out for the wind speed,wind direction angle and wind temperature respectively,and it was concluded that the difference in surface temperature of potatoes and impurities was nonlinear and positively related to the wind speed,wind direction Angle and wind temperature in a certain period of time.Finally,potatoes and impurities were classified on the test platform.The experimental results showed that the optimal image recognition conditions were wind speed of 4 m/s,wind temperature of 40℃and wind direction angle of 90°,and the recognition success rate was 97%.It can provide technical reference for potato removal in harvesting-storage process.

关 键 词：马铃薯 热激励 热红外成像 杂质剔除 

分 类 号：S226.5[农业科学—农业机械化工程]