基于嵌入式的塑料瓶喉口缺陷检测系统研究

发布时间：2018-04-21 02:11

本文选题：DM6446 + 嵌入式　；参考：《华侨大学》2017年硕士论文

【摘要】：塑料瓶包装行业一直以来都是我国的一大基础产业,塑料瓶被广泛应用于食品、化妆品、洗涤剂、医药品、以及工业药品、工业容器等领域。平时人们所用生活用品的包装容器大多以塑料瓶的形式出现,且数量巨大。然而现实生产过程中由于制造设备的局限,塑料瓶喉口易发生堵料现象。目前,一般都采用人工检测,在快速的生产流水线环境中很难保证产品质量稳定性和产出效率,同时也增加了生产成本。因此,本文针对塑料瓶喉口缺陷检测系统的研发具有很重要的现实意义。本文首先介绍了机器视觉技术的发展概况和应用前景,并进一步分析了嵌入式系统与其结合的优势特点。针对常用的塑料瓶喉口缺陷的检测现状,本文基于达芬奇技术的DM6446视频图像处理平台,采用机器视觉检测的方法,提出了将其应用于塑料瓶喉口缺陷检测的系统方案。然后,分析塑料瓶喉口缺陷产生的原因以及现有的主流机器视觉处理平台,提出并设计该检测系统的总体方案。其中包括具体硬件处理平台的选择以及该系统的总体结构方案和工作流程。在硬件方面,完成了对主要芯片的选取及介绍和对电源模块、存储模块、视频输入输出模块、报警与动作控制模块等的硬件原理设计。以该硬件平台为基础搭建和配置系统软件开发环境,完成了ARM子系统对解码芯片、视频处理子系统以及其它设备的参数配置。在软件算法方面,对系统软件流程进行了设计并研究了DSP子系统中对图像处理的一些算法,其中主要包括图像灰度化、图像增强、图像滤波、自适应二值化、边缘检测、缺陷检测以及相应的算法优化等。最后,在搭建好的硬件平台上,基本实现了塑料瓶喉口缺陷的检测。通过实验证明,该系统可以实现生产线上塑料瓶喉口缺陷的机器视觉检测,代替主流的人工检测,提高流水线生产的自动化程度。本论文的主要创新点在于基于DM6446的图像处理开发平台,将嵌入式技术和机器视觉技术结合并应用到塑料瓶喉口缺陷的检测中,以及在检测过程中对相关算法的优化创新。本文的系统研究可以大幅度提高相关产品的质量检测效率和在线生产的自动化程度,对于以后研发高性价比、稳定性好、体积小的高度自动化在线生产质量检测设备,具有工程实用意义和参考价值。
[Abstract]:Plastic bottle packaging industry has always been a major basic industry in our country. Plastic bottles are widely used in food, cosmetics, detergent, pharmaceutical products, industrial drugs, industrial containers and other fields. Usually, the packaging containers for people's daily articles are mostly in the form of plastic bottles, and the quantity is huge. However, the actual production process is made up of the production process. At present, artificial detection is commonly used in the throat of plastic bottles. It is difficult to ensure product quality stability and production efficiency in the rapid production line environment, but also increases production cost. Therefore, this paper has an important reality for the research and development of the throat defect detection system for plastic bottles. This paper first introduces the development and application prospects of machine vision technology, and further analyzes the advantages and characteristics of the embedded system and its combination. In view of the current detection status of common plastic bottle throat defects, this paper, based on the DM6446 video image processing platform based on the Da Vinci technology, uses the method of machine vision detection. The system is applied to the detection of the throat defects of plastic bottles. Then, the causes of the defects in the throat of the plastic bottles and the existing main machine vision processing platform are analyzed. The overall scheme of the detection system is proposed and designed, including the selection of the specific hardware processing platform, the overall structure scheme and the working process of the system. On the hardware aspect, we have completed the selection and introduction of the main chip and the hardware design of the power module, the storage module, the video input and output module, the alarm and the action control module. Based on the hardware platform, the system software development environment is built and configured, and the decoding chip, video processing subsystem and the ARM subsystem are completed. In the aspect of software algorithm, the software process is designed and some algorithms for image processing in DSP subsystem are studied, including image grayscale, image enhancement, image filtering, adaptive two value, edge detection, defect detection and corresponding algorithm optimization. Finally, it is built well On the hardware platform, the flaw detection of the plastic bottle throat is basically realized. Through the experiment, it is proved that the system can realize the machine vision inspection of the plastic bottle throat defect on the production line, instead of the mainstream manual detection, and improve the automation degree of the production line. The main creation point of this paper is the DM6446 based image processing and development platform. The embedded technology and machine vision technology are combined and applied to the detection of plastic bottle throat defects, as well as the optimization of the related algorithms in the process of detection. The system research in this paper can greatly improve the quality detection efficiency of the related products and the automation range of online production, and have high performance price ratio and good stability for future research and development. The small size highly automated on-line production quality testing equipment has practical significance and reference value.

【学位授予单位】：华侨大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ320.72;TP391.41

【相似文献】