基于机器视觉的禽蛋分级分选机器人

发布时间：2018-02-04 05:42

本文关键词： 禽蛋机器视觉分级分选机器人　出处：《宁夏大学》2017年硕士论文　论文类型：学位论文

【摘要】：中国是一个农业大国,禽蛋产业一直存在着广阔的发展前景。随着人们生活水平的提高,对禽蛋的品质要求也越来越高,禽蛋个体适中、表面洁净的禽蛋,往往是消费者首选的对象。就我国而言,禽蛋分级大多采用人工分级的方式,这样不仅成本高、效率低、人为因素影响大,而且食品的安全性也存在着威胁。国内的一些分级设备的分级标准过于单一,国外的设备售价又较高。随着科学技术的发展,越来越多的新技术应用到实际生产环节中,为生产带来了便利。本文探索利用机器视觉技术,对禽蛋的大小、壳色、表面污染程度进行检测算法研究,利用传送带装置、真空吸盘,结合三角洲式并联机械臂,实现对禽蛋的运输、抓取、分级任务。1.搭建图像采集系统,确立图像采集相机、光源以及采集背景等,为获取良好的禽蛋原始图像奠定良好的基础。2.上位机核心处理器采用BCM2837的64位ARMv8四核处理器,利用数字图像处理技术,对采集回的禽蛋图像进行预处理,其中对图像灰度化、图像增强、去噪、二值化等采取了多种处理方式,根据实际效果,选择最佳处理方式。3.对选择的禽蛋大小、壳色及污染程度三个指标进行算法研究,实现禽蛋相应特征值的提取,并设立评估体系,对检测算法性能进行评估。4.下位机以ATmega2560核心处理器的单片机,控制电磁阀实现真空吸盘抓取禽蛋、并联机械臂对禽蛋搬运分级动作。实验结果表明,本文研究的一种禽蛋分级分选机器人,能够较好的实现禽蛋的分级分选任务,并且禽蛋的分级效果也较明显。
[Abstract]:China is a large agricultural country, the poultry egg industry has a broad development prospects. With the improvement of people's living standards, the quality of eggs is higher and higher. As far as our country is concerned, poultry egg grading mostly adopts artificial grading, which is not only high cost, low efficiency, but also has great influence on human factors. There is also a threat to the safety of food. The classification standard of some domestic grading equipment is too single, and the price of foreign equipment is higher. With the development of science and technology. More and more new technologies are applied to the actual production, which bring convenience to production. This paper explores the use of machine vision technology to detect the size of eggs, shell color, surface pollution degree of detection algorithm. Using conveyor belt device, vacuum sucker, combined with delta parallel manipulator, to achieve the transport of eggs, grab, grading task. 1. Build an image acquisition system, establish an image acquisition camera. Light source and collecting background, which lay a good foundation for obtaining good original image of poultry eggs. The upper computer core processor uses BCM2837 64-bit ARMv8 quad-core processor. Using the digital image processing technology, the collected egg image is preprocessed. Among them, the image grayscale, image enhancement, denoising, binarization and so on are processed in a variety of ways, according to the actual results. Choose the best treatment method .3.Research the algorithm of the three indexes of egg size, shell color and pollution degree, realize the extraction of the corresponding characteristic value of poultry egg, and set up the evaluation system. Evaluation of the performance of the detection algorithm .4.The lower computer with ATmega2560 core processor of the single-chip microcomputer, control solenoid valve to realize the vacuum sucker to grab eggs. The experiment results show that a kind of egg grading robot studied in this paper can achieve the classification and sorting task of poultry eggs, and the classification effect of eggs is obvious.
【学位授予单位】：宁夏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TP242

【参考文献】