基于图像处理的透明胶囊缺陷检测系统研究

发布时间：2018-02-01 10:33

本文关键词： Canny 透明胶囊开运算缺陷检测直方图均衡化　出处：《中国海洋大学》2015年硕士论文　论文类型：学位论文

【摘要】：随着计算机的广泛应用,信息技术开始快速发展,并在工业领域逐渐占据重要地位。与此同时,随着数字图像处理技术的逐步完善,其广阔的适用性、准确性和高效性,使其在工业领域比重越来越大。随着数字图像处理技术的重要性日益凸显,采用数字图像处理技术来节省人力成本和减少失误逐渐成为必然。利用数字图像对物体表面进行缺陷检测是计算机视觉和数字图像处理应用的重要组成部分,其主要思路是找到图像中的分割边缘,并根据边缘及内部特性来判断物体表面是否存在缺陷。利用数字图像处理技术代替人工识别进行缺陷检测,不仅具有高效性和稳定性,更能够排除人工失误等造成的干扰。透明胶囊缺陷检测主要包括长度、黑点、气泡、破洞、切口和夹皱等方面,目前国内尚无完善的透明胶囊检测系统。本文目标是设计出能够准确、高效检测透明胶囊缺陷的检测系统,采用的设计方案是：对传输中的胶囊通过工业用相机提取图像,针对不同缺陷利用膨胀、腐蚀、直方图均衡化等不同技术进行初步处理,然后利用改进Canny算子提取二值化边缘图像,并对其进行不同的缺陷判断。对于不合格胶囊,机器将在传输过程中对其进行分离,保证胶囊的质量,并统计合格和不合格胶囊的数量。论文首先讲述了国内外研究现状,对使用的检测技术和开发环境进行了描述；然后分别针对几种代表性的缺陷使用的算法进行了说明；其次,展示了系统的总体设计、实现与胶囊检测结果并进行了分析；最后,对全文进行了总结,提出了可能的优化方案,并对系统的扩展进行了展望。
[Abstract]:With the wide application of computer, information technology has developed rapidly and occupied an important position in the field of industry. At the same time, with the gradual improvement of digital image processing technology, its wide applicability. Accuracy and efficiency make it more and more important in the industrial field. With the increasing importance of digital image processing technology. It is inevitable to use digital image processing technology to save manpower cost and reduce errors. It is an important part of computer vision and digital image processing applications to use digital image to detect object surface defects. The main idea is to find the segmentation edge in the image and determine whether there are defects on the surface of the object according to the edge and internal characteristics. The digital image processing technology is used instead of manual recognition to detect the defects. It not only has high efficiency and stability, but also can eliminate the interference caused by human error. Transparent capsule defect detection mainly includes length, black spot, bubble, hole, incision and wrinkle. At present, there is no perfect transparent capsule detection system in China. The purpose of this paper is to design an accurate and efficient detection system for transparent capsule defects. The design scheme adopted is: the capsule in transmission is extracted by industrial camera, and the different technologies such as expansion, corrosion and histogram equalization are used to deal with the different defects. Then the improved Canny operator is used to extract the binary edge image, and different defects are judged. For the unqualified capsules, the machine will separate them in the transmission process to ensure the quality of the capsules. And count the number of qualified and unqualified capsules. Firstly, the paper describes the current situation of research at home and abroad, and describes the detection technology and development environment. Then, the algorithms used in several representative defects are explained. Secondly, the overall design of the system is presented, and the test results of the capsule are realized and analyzed. Finally, the paper summarizes the whole paper, puts forward the possible optimization scheme, and looks forward to the expansion of the system.
【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41

【相似文献】