【作者】 刘荣；

【导师】 李久熙； 王学静；

【作者基本信息】 河北农业大学 ， 工程硕士（专业学位）， 2021， 硕士

【摘要】 暗斑是蛋壳质量评价指标之一,蛋壳暗斑不仅严重损害了鸡蛋的外观品质,而且细菌及微生物在暗斑处更容易侵入鸡蛋内容物从而影响鸡蛋的存储时间,增加食用鸡蛋的安全隐患。自然光照条件下蛋壳表面的暗斑难以检测和度量,本文设计了一种小型化暗斑蛋并行处理的自动检测和分拣装置,由上料装置、转运装置、翻转装置、图像采集装置、暗斑检测与度量模块组成。该装置通过流水线方式自动输蛋、翻转鸡蛋,采集完整蛋壳表面图像、进行暗斑数量检测,实现暗斑的自动化识别与度量,并根据度量结果进行蛋品分拣。实验结果表明,该装置能够快速、准确实现鸡蛋图像自动采集、蛋壳暗斑分布数量和投影面积占比计算。该装置克服了人工标注主观性强的问题,提高了检测精度;暗斑蛋检测及分拣的处理速度为1765枚/h。该研究结果为高通量的鸡蛋品质在线检测提供了自动化手段。本文的研究内容主要包括:(1)在需求分析基础上对暗斑检测及分拣装置进行整体设计,对整机总体结构、工作流程、总体布局进行设计,并对鸡蛋暗斑检测过程中的关键结构如输送装置、翻转装置进行方案优选。(2)根据鸡蛋的运动特性及尺寸参数对上料装置、转运装置、翻转装置进行设计;并通过鸡蛋尺寸测量试验、翻转装置测试试验验证装置工作的可靠性。上料装置通过微型电推杆与电磁铁实现间歇输蛋;转运装置利用步进电机实现间歇运动;翻转装置为二自由度机械爪,具有一个转动关节和两个执行机构。(3)利用ADAMS软件对装置进行了运行仿真。验证了在暗斑图像处理时间不定的条件下,翻转装置与转运装置能够实现暗斑检测及分拣。(4)根据暗斑的特性,参考基于图像的鸡蛋裂纹检测方法和人工检测暗斑所遇到的问题,设计出鸡蛋暗斑检测的图像采集装置,并设计了适合鸡蛋暗斑图像的采集暗箱。(5)进行鸡蛋暗斑检测及分拣装置样机试制与试验,解决装配中遇到的问题。随机选取的300枚鸡蛋进行人工组和装置组的暗斑识别与分拣结果数据对比,试验结果表明装置检测的准确率为96.67%,比人工高2.67%;装置检测并分拣的平均速度为3.1秒/个,较人工平均速度快。鸡蛋的暗斑检测及分拣装置有效解决了鸡蛋的暗斑检测和分拣工作中人工耗费大的问题,实现了暗斑蛋检测及分拣的自动化,提高检测及分拣的准确性,促进了鸡蛋暗斑相关研究的进展,为鸡蛋生产企业提供了更高效的鸡蛋暗斑检测及分拣方式。更多还原

【Abstract】 Eggshell dark spot are one of the evaluation indicators of Eggshell quality.Eggshell dark spot not only seriously damage the appearance quality of eggs,but also bacteria and microorganisms are more likely to invade the egg content at the Eggshell,which affects the storage time of eggs and increases the safety of eating eggs.Hidden dangers.It is difficult to detect and measure the Eggshell dark spot on the Eggshell surface under natural light conditions.This paper designs a miniaturized automatic detection and sorting device for parallel processing of Eggshell eggs.It consists of a loading device,a transfer device,a turning device,an image acquisition device,Spot detection and measurement module,sorting device.The device automatically transfers eggs,flips eggs through an assembly line,collects complete Eggshell surface images,detects the number of dark spot,realizes automatic identification and measurement of dark spot,and sorts eggs according to the measurement results.The experimental results show that the device can quickly and accurately realize the automatic collection of egg images,the calculation of the number of Eggshell dark spot and the proportion of projected area.The device overcomes the problem of strong subjectivity of manual labeling and improves detection accuracy;the processing speed of dark spot egg detection and sorting is 1765 eggs/h.The results of this study provide an automated means for high-throughput online detection of egg quality.The research content of this article mainly includes:(1)On the basis of demand analysis,carry out the overall design of the dark spot detection and sorting device,design the overall structure,work flow,and overall layout of the whole machine,and design the key structures in the egg dark spot detection process such as conveying device and flip The device is optimized.(2)According to the movement characteristics and size parameters of the eggs,the feeding device,the transfer device,and the turning device are designed;and the reliability of the device is verified through the egg size measurement test and the turning device test test.The feeding device realizes intermittent egg delivery through miniature electric push rods and electromagnets;the transport device uses stepping motors to realize intermittent movement;the turning device is a two-degree-of-freedom manipulator with a rotating joint and two actuators.(3)Using ADAMS software to simulate the operation of the device.It is verified that the turning device and the transfer device can realize the detection and sorting of the dark spot under the condition of uncertain image processing time.(4)According to the characteristics of dark spot,referring to the image-based egg crack detection method and the problems encountered in manual detection of dark spot,an image acquisition device for egg dark spot detection is designed,and a dark box suitable for egg dark spot images is designed..(5)Carry out trial production and experiment of egg dark spot detection and sorting device prototype to solve the problems encountered in assembly.300 eggs were randomly selected to compare the results of dark spot recognition and sorting between the manual group and the device group.The test results show that the accuracy of the device detection is 96.67%,which is 2.67%higher than the manual.The average speed of detection and sorting is 2.04 seconds/piece,twice the speed of manual work.The egg dark spot detection and sorting device effectively solves the problem of large manual labor in the dark spot detection and sorting of eggs,realizes the automation of dark spot egg detection and sorting,improves the accuracy of detection and sorting,and promotes The development of related research on egg dark spot has provided egg manufacturers with more efficient egg dark spot detection and sorting methods.更多还原