医药大输液可见异物的视觉检测机器人技术研究
|
摘要
医药大输液是我国医药行业五大重要制剂之一，是医疗机构日常必须使用的药品，在现代临床上占据十分重要的地位。但是，由于生产工艺以及封装技术的原因，在灌装过程中，大输液产品中可能含有玻璃屑、纤维、毛发、漂浮物等可见异物，异物的存在将危害用药安全。目前，我国制药企业由于缺乏相关配套技术和生产成本考虑，大部分都没有在瓶装液体药品生产线中引进或建立自动化医药检测装备系统，而基本上都是采用人工灯检的方法，要求工作人员在暗室中进行，配备简单的检测灯箱，通过目视检测药液中的可见异物，这种方法检查速度慢、操作繁琐、可靠性差，还容易对药品造成二次污染，从而危害制药安全。而基于机器视觉的药品检测机器人能够实现药品在线、高速、高精度的自动化检测。我国在灌装后的药品视觉检测设备研究开发方面远远落后于国外发达国家，而这与我国大输液生产使用量世界首位极不相称。
为此，针对现有人工检测方法存在的问题和我国药企的需求，本论文对大输液智能视觉检测机器人技术进行了深入的研究。论文介绍了大输液视觉检测的概念和要求，从大输液视觉检测机器人的检测原理、机械结构、电气控制系统、光学照明与图像获取方案到可见异物的图像检测算法都进行了详细研究，在此基础上，开发出视觉检测机器人并进行了多项指标测试和验证。
本论文的研究工作、主要成果和创新点包括以下几个方面：
1、介绍了医药大输液视觉检测机器人的研究背景与意义，分析了我国医药市场和制药装备的现状，以及大输液药品的生产制造流程，指出在医药检测环节人工灯检存在的问题；概述了医药视觉检测机器人相关的一些重要机器视觉技术，包括光源照明技术、视觉成像等，还介绍了机器视觉在医药、饮料、电子生产线检测中的应用情况；最后分析并总结了国外开发类似检测技术与设备的相关成果。
2、根据医药生产企业对大输液药品在线检测的应用要求，结合我国大输液的实际生产工艺，分析了医药自动化生产线上智能检测机器人的技术可行性。论文提出了模拟人工检测动作，药品高速旋转-急停-跟踪拍摄的视觉检测原理。设计了用于检测100ml以下大输液的机器人机械结构和检测100ml及以上大输液的机器人机械结构及多个重要机械机构，包括：抓瓶机构（药瓶夹持机构）、旋转搓瓶机构、次品剔除机构等；设计了由控制子系统和图像处理子系统两部分组成的总体控制系统。在控制子系统中，开发了基于多PC机并行处理的分布式控制结构，设计了系统中的旋转-急停-跟踪拍摄的控制时序、多段分区域搓瓶时间控制、以及跟踪摆臂运动控制等。在图像采集与处理子系统，从相机镜头选型到光学系统设计，经过反复实验，实现了在高速运动中获取清晰的药品图像。论文还分析了医药大输液视觉检测机器人的系统工作过程及软件构架方案。
3、提出了一种基于FFT频域变换的图像校正与配准方法，测试了多组检测序列图像，完成对序列图像参数的粗略配准，在此基础上，采用Powell最优化搜索算法进一步寻找最佳匹配参数，以图像子集的平均互相关度量值为配准的代价函数，最终获得图像的精确配准参数，配准精度达到了亚像素级。为了减少后续处理的运算量，提出一种基于概率统计的方法对输液图像进行感兴趣区域提取，首先对获得的图像进行直方图均衡化，增强图像整体对比度，然后对药品图像进行二维最大熵阀值分割，最后使用概率统计的方法求出图像中包含药液区域的左右和上下边界位置。为了抑制背景噪声的干扰，提出了一种基于极值的自适应均值滤波算法，与传统的滤波算法比较，该算法能够去除噪声的同时较好地保留边缘等细节信息，降低图像处理后的模糊化程度，可以有效的对检测区域内的药液图像进行了噪声滤除。
4、结合药液中可见异物的检测特点，提出基于细胞神经网络（Cellular neuralnetworks, CNN）的药液图像分割方法，设计恰当的细胞神经网络模板分割图像，然后为改变CNN的线性加权联结方式，引入了非线性模糊运算min/max联结权，设计了模糊细胞神经网络（FCNN）结构，通过实验测试，FCNN分割效果优于传统的CNN，但在边缘检出上其效果尚不理想；本文针对此问题，提出改进型模糊细胞神经网络（IFCNN），研究了其收敛性和稳定性，实验测试结果表明，它能有效的克服现有方法无法解决的边缘检出问题，使得异物分割图像能更好的接近真实图像。
5、根据药液异物图像形态复杂、类型多样等特点，对连续多帧医药图像选取了目标的一系列形态特征、统计特征、运动特征等，介绍了使用的各个特征参数的计算方法，然后针对实验图像，提取各个目标的特征参数，并进行分析。在此基础上，提出了一种改进的ReliefF算法-k个最近邻ReliefF算法来选取特征，通过该特征选择算法滤除无关特征，该算法能在进行权值迭代中用k个最近邻距离的平均代替传统的一个最近邻，从而较大程度的减少了噪声等虚假特征对权值的影响，使得特征的选择更为精确。论文分析了支持向量机和Boosting之后，通过实验比较两者在药液中多种异物分类应用中的优缺点，提出了基于SVM的AdaBoosting多值分类算法，并给出实验结果，该方法集成在药品缺陷检测软件系统中，取得了良好的效果。
6、研制出了一台大输液检测机器人样机和一套软件平台。从工程应用角度，详细介绍了研制的大输液检测机器人硬件系统，对进瓶机构、抓瓶与搓瓶机构以及电气控制等多个组成部分进行了说明，并开发了检测与分析软件各组成模块，在系统性能的测试过程中，通过国际通用测试方法（Knapp-Kushner）对本文设计的药液异物检测系统和算法的有效性进行验证，另一方面也对检测方法的重复性、检测精度、不同种类异物检测性能进行了测试，完全能够满足输液在线检测的要求。
本论文通过理论分析研究和实验证明了提出的大输液医药视觉检测机器人光学、机械、电气结构的合理性，以及大输液可见异物检测方法的有效性和可行性，研制的检测机器人系统做为国家863课题―大型高速医药自动化生产线上的产品检测包装智能机器人‖的重要成果之一，在2011.12.9通过了国家科技部组织的专家验收。检测机器人现场应用表明，本文研制的检测系统解决了实际应用的大部分问题，将在医药自动化检测中发挥重要的作用，具有极高的实用价值和推广应用前景。
Medical infusion is one of the five important preparations of the pharmaceuticalindustry in China. They are frequently-used drugs in medical institutions and play avery important role in modern clinic. However, some visible foreign substances mayappear in the bottled infusion during the processes of producing, filling and packaging.These foreign substances will turn out to be a serious threat to patients. At present,due to the lack of pivotal technologies and for cost reduction, most of the Chinesepharmaceutical manufacturers adopt the traditional manual inspection, instead ofbringing in the automatic medical inspection equipments in most of the bottled liquidpharmaceutical production lines. In the manual inspection, trained workers check thevisible foreign substances in bottled liquid in front of the light boxes in darkrooms.Obviously, the manual inspection is slow, complex and unreliable, and usually bringsin extra pollution to the drugs, thus is unsafe. Fortunately, vision based medicineinspection machine can achieve online, high speed and high accuracy, and automaticdetection. The research of vision based medicine inspection equipment for bottledinfusion in China is far behind the developed countries, which is vastlydisproportionate with the top-ranking Chinese infusion usage.
Hence, to solve the existing problems in manual inspection and meet therequirements of pharmaceutical companies, this dissertation carried out a deepresearch on the intelligent vision based infusion inspection technologies. Theconcepts and requirements of vision based inspection of bottled infusion areintroduced firstly. Then the detection principle, mechanical structure, electricalcontrol system, optical illumination, image capturing system and detection algorithmsfor visible foreign substances are studied in detail. Accordingly, a sample visualinspection machine is developed, on which a series of indicators are tested andvalidated.
Major achievements and innovations of this dissertation are listed below:
1. Research background and significance of the visual infusion inspectionmachine are introduced. Status of Chinese pharmaceutical market and equipment arealso described. Then the producting process of the bottled infusion is introduced andthe problems in the manual inspection are pointed out. Meanwhile, some typicalmachine vision technologies of inspection machine are overviewed, such as light sources and lighting technology, visual imaging, as well as their applications in thearea of pharmaceutical, beverage, electronics and production line inspection. Finally,foreign similar development results of detection technologies and equipments areanalyzed and summarized.
2. According to the requirements of pharmaceutical manufacturers’ onlineinspection, intelligent inspection machine’s technical feasibilities are analyzed withthe practical producting process. To simulate the manual inspection, a strategy--―high-speed rotation of drugs, emergency stop, tracking and image capturing‖isproposed. Meanwhile, a series of essential mechanical structures are designed, such asbottle grasper (bottle holding mechanism), rotating and twisting mechanism, defectiverejection mechanism,etc. The overall system consists of the control subsystem and theimage processing subsystem. In the control subsystem, a distributed control structurebased on parallel processing of multi-IPC is developed. The timing of rotation,braking, and tracking and snapshotting, the time control strategy of sub-regionalbottle rubbing in the multi-stage is proposed. In addition, the control method of thetracking arm motion is designed. In the image acquisition and processing subsystem,legible images of bottled infusion are acquired after plenty of repeated experimentsfrom the selection of the camera lens to the optical system design. In addition, theworking process and the software architecture of the visual inspection robot systemfor pharmaceutical infusion are analyzed in detail.
3. An FFT frequency domain transform based image calibration and registrationmethod is proposed, testing a series of sequence images to complete the roughalignment of the sequence image parameters. On this basis, through Powelloptimization searching algorithms, the best matching parameters are w orked out.Considering the average cross-relation value of image subset as the cost function ofmatching, the precise image registration parameters are obtained, which achieves thealignment accuracy of the sub-pixel level. In order to reduce the computationalcomplexity of further processing, a method based on probability and statistics for theinterest area extraction of infusion image is proposed. Firstly, the histogramequalization operation is applied to the images to enhance the overall image contrast.Then, the2D maximum entropy threshold segmentation is implemented. Finally,through the method of probability and statistics, the position of the image containingthe liquid region is worked out. In order to suppress the interference of backgroundnoise, an adaptive mean filtering algorithm based on maximum value is proposed.Compared with the traditional filtering algorithms, our algorithm can remove noise successfully while keeping the edge and other details in the images, and relaxing thefuzziness, which can effectively filter noise within the detecting area of the infusionimage.
4. Combined with the characteristics of the inspection for visible foreignsubstances of infusions, an infusion image segmentation algorithm based on CellularNeural Networks (CNN) is proposed. An appropriate CNN template is designed tochange the CNN weighted linear links. Meanwhile, the min/max linking weights ofnonlinear fuzzy operation is introduced, and the structure of Fuzzy Cellular NeuralNetworks (FCNN) is designed. Experiments showed that FCNN was more effectivethan traditional CNN, but FCNN's effect on edge detection was not so perfect. Tosolve this problem, an Improved Fuzzy Cellular Neural Network (IFCNN) is proposedand its convergence and stability are studied. Experimental results showed thatIFCNN can effectively solve the problem appeared in the edge detection that may notbe solved by existing methods, which achieved better approximation to the originalimages.
5. According to the features of complexity and multi-character of the foreignsubstances, a series of morphological, statistical and motion features are selectedfrom the continuous medical image sequences. The calculation method of every usedfeature is introduced. The feature parameters of every target in the experimentalimages are extracted and analyzed. Accordingly, an improved ReliefF algorithm withk nearest neighborhoods is proposed to extract the features. By filtering out irrelevantfeatures with the feature selection algorithm, the algorithm can replace the traditionalnearest neighbor value with the average value of distance to k nearest neighborhoods,to reduce effects of false features such as noises on weight values vastly, and to makefeature extraction more precisely. After analyzing SVM and Boosting algorithms,experiments are carried out to compare the advantages and disadvantages of the twomethods in the classification applications for a variety of foreign substances. Then theAdaBoosting multi-class classification algorithm based on SVM is proposed, andtested by related experiments. In addition, the algorithm is integrated in the softwareof the pharmaceutical defedct inspection line that achieves great outcomes.
6. A sample bottled infusion inspection machine and the related softwareplatform are developed. From the point of engineering application, the softwaresystem of bottled infusion inspection machine is described in detail, ranging fromentrance structure, gripper, rubbing device to electrical control modules, etc. Thedetection and analysis software modules are developed. In the process of the system performance testing, the effectiveness of the bottled infusion inspection machine andrelated algorithms are validated with international verification method (Knapp-Kushner Test). On the other hand, the repeatability, detection accuracy and theperformance on detecting different types of foreign substances of the inspectionmethod are also tested, whose results show that sample machine can meet therequirements of the bottled infusion online inspection effectively.
In this dissertation, through theoretical analysis and experiments, the validityand feasibility of the optical, mechanical, electrical structures and inspectionalgorithms in the proposed bottled infusion visual inspection machine are proved. Asone of the important outcomes of the National High Technology Research andDevelopment Program of China "Intelligent Inspecting and Packaging Robot inLarge-Scale High-Speed Pharmaceutical Automatic Production Line", the develo pedinspection machine is accepted by the experts of the Ministry of Science andTechnology on December9th,2011. The field applications of the inspection machineshow that the developed system can solve most problems in the field applications, andwill play an important role in the automatic detection of pharmaceutical industry, andhave a promising future in a wide range of applications.
为此，针对现有人工检测方法存在的问题和我国药企的需求，本论文对大输液智能视觉检测机器人技术进行了深入的研究。论文介绍了大输液视觉检测的概念和要求，从大输液视觉检测机器人的检测原理、机械结构、电气控制系统、光学照明与图像获取方案到可见异物的图像检测算法都进行了详细研究，在此基础上，开发出视觉检测机器人并进行了多项指标测试和验证。
本论文的研究工作、主要成果和创新点包括以下几个方面：
1、介绍了医药大输液视觉检测机器人的研究背景与意义，分析了我国医药市场和制药装备的现状，以及大输液药品的生产制造流程，指出在医药检测环节人工灯检存在的问题；概述了医药视觉检测机器人相关的一些重要机器视觉技术，包括光源照明技术、视觉成像等，还介绍了机器视觉在医药、饮料、电子生产线检测中的应用情况；最后分析并总结了国外开发类似检测技术与设备的相关成果。
2、根据医药生产企业对大输液药品在线检测的应用要求，结合我国大输液的实际生产工艺，分析了医药自动化生产线上智能检测机器人的技术可行性。论文提出了模拟人工检测动作，药品高速旋转-急停-跟踪拍摄的视觉检测原理。设计了用于检测100ml以下大输液的机器人机械结构和检测100ml及以上大输液的机器人机械结构及多个重要机械机构，包括：抓瓶机构（药瓶夹持机构）、旋转搓瓶机构、次品剔除机构等；设计了由控制子系统和图像处理子系统两部分组成的总体控制系统。在控制子系统中，开发了基于多PC机并行处理的分布式控制结构，设计了系统中的旋转-急停-跟踪拍摄的控制时序、多段分区域搓瓶时间控制、以及跟踪摆臂运动控制等。在图像采集与处理子系统，从相机镜头选型到光学系统设计，经过反复实验，实现了在高速运动中获取清晰的药品图像。论文还分析了医药大输液视觉检测机器人的系统工作过程及软件构架方案。
3、提出了一种基于FFT频域变换的图像校正与配准方法，测试了多组检测序列图像，完成对序列图像参数的粗略配准，在此基础上，采用Powell最优化搜索算法进一步寻找最佳匹配参数，以图像子集的平均互相关度量值为配准的代价函数，最终获得图像的精确配准参数，配准精度达到了亚像素级。为了减少后续处理的运算量，提出一种基于概率统计的方法对输液图像进行感兴趣区域提取，首先对获得的图像进行直方图均衡化，增强图像整体对比度，然后对药品图像进行二维最大熵阀值分割，最后使用概率统计的方法求出图像中包含药液区域的左右和上下边界位置。为了抑制背景噪声的干扰，提出了一种基于极值的自适应均值滤波算法，与传统的滤波算法比较，该算法能够去除噪声的同时较好地保留边缘等细节信息，降低图像处理后的模糊化程度，可以有效的对检测区域内的药液图像进行了噪声滤除。
4、结合药液中可见异物的检测特点，提出基于细胞神经网络（Cellular neuralnetworks, CNN）的药液图像分割方法，设计恰当的细胞神经网络模板分割图像，然后为改变CNN的线性加权联结方式，引入了非线性模糊运算min/max联结权，设计了模糊细胞神经网络（FCNN）结构，通过实验测试，FCNN分割效果优于传统的CNN，但在边缘检出上其效果尚不理想；本文针对此问题，提出改进型模糊细胞神经网络（IFCNN），研究了其收敛性和稳定性，实验测试结果表明，它能有效的克服现有方法无法解决的边缘检出问题，使得异物分割图像能更好的接近真实图像。
5、根据药液异物图像形态复杂、类型多样等特点，对连续多帧医药图像选取了目标的一系列形态特征、统计特征、运动特征等，介绍了使用的各个特征参数的计算方法，然后针对实验图像，提取各个目标的特征参数，并进行分析。在此基础上，提出了一种改进的ReliefF算法-k个最近邻ReliefF算法来选取特征，通过该特征选择算法滤除无关特征，该算法能在进行权值迭代中用k个最近邻距离的平均代替传统的一个最近邻，从而较大程度的减少了噪声等虚假特征对权值的影响，使得特征的选择更为精确。论文分析了支持向量机和Boosting之后，通过实验比较两者在药液中多种异物分类应用中的优缺点，提出了基于SVM的AdaBoosting多值分类算法，并给出实验结果，该方法集成在药品缺陷检测软件系统中，取得了良好的效果。
6、研制出了一台大输液检测机器人样机和一套软件平台。从工程应用角度，详细介绍了研制的大输液检测机器人硬件系统，对进瓶机构、抓瓶与搓瓶机构以及电气控制等多个组成部分进行了说明，并开发了检测与分析软件各组成模块，在系统性能的测试过程中，通过国际通用测试方法（Knapp-Kushner）对本文设计的药液异物检测系统和算法的有效性进行验证，另一方面也对检测方法的重复性、检测精度、不同种类异物检测性能进行了测试，完全能够满足输液在线检测的要求。
本论文通过理论分析研究和实验证明了提出的大输液医药视觉检测机器人光学、机械、电气结构的合理性，以及大输液可见异物检测方法的有效性和可行性，研制的检测机器人系统做为国家863课题―大型高速医药自动化生产线上的产品检测包装智能机器人‖的重要成果之一，在2011.12.9通过了国家科技部组织的专家验收。检测机器人现场应用表明，本文研制的检测系统解决了实际应用的大部分问题，将在医药自动化检测中发挥重要的作用，具有极高的实用价值和推广应用前景。
Medical infusion is one of the five important preparations of the pharmaceuticalindustry in China. They are frequently-used drugs in medical institutions and play avery important role in modern clinic. However, some visible foreign substances mayappear in the bottled infusion during the processes of producing, filling and packaging.These foreign substances will turn out to be a serious threat to patients. At present,due to the lack of pivotal technologies and for cost reduction, most of the Chinesepharmaceutical manufacturers adopt the traditional manual inspection, instead ofbringing in the automatic medical inspection equipments in most of the bottled liquidpharmaceutical production lines. In the manual inspection, trained workers check thevisible foreign substances in bottled liquid in front of the light boxes in darkrooms.Obviously, the manual inspection is slow, complex and unreliable, and usually bringsin extra pollution to the drugs, thus is unsafe. Fortunately, vision based medicineinspection machine can achieve online, high speed and high accuracy, and automaticdetection. The research of vision based medicine inspection equipment for bottledinfusion in China is far behind the developed countries, which is vastlydisproportionate with the top-ranking Chinese infusion usage.
Hence, to solve the existing problems in manual inspection and meet therequirements of pharmaceutical companies, this dissertation carried out a deepresearch on the intelligent vision based infusion inspection technologies. Theconcepts and requirements of vision based inspection of bottled infusion areintroduced firstly. Then the detection principle, mechanical structure, electricalcontrol system, optical illumination, image capturing system and detection algorithmsfor visible foreign substances are studied in detail. Accordingly, a sample visualinspection machine is developed, on which a series of indicators are tested andvalidated.
Major achievements and innovations of this dissertation are listed below:
1. Research background and significance of the visual infusion inspectionmachine are introduced. Status of Chinese pharmaceutical market and equipment arealso described. Then the producting process of the bottled infusion is introduced andthe problems in the manual inspection are pointed out. Meanwhile, some typicalmachine vision technologies of inspection machine are overviewed, such as light sources and lighting technology, visual imaging, as well as their applications in thearea of pharmaceutical, beverage, electronics and production line inspection. Finally,foreign similar development results of detection technologies and equipments areanalyzed and summarized.
2. According to the requirements of pharmaceutical manufacturers’ onlineinspection, intelligent inspection machine’s technical feasibilities are analyzed withthe practical producting process. To simulate the manual inspection, a strategy--―high-speed rotation of drugs, emergency stop, tracking and image capturing‖isproposed. Meanwhile, a series of essential mechanical structures are designed, such asbottle grasper (bottle holding mechanism), rotating and twisting mechanism, defectiverejection mechanism,etc. The overall system consists of the control subsystem and theimage processing subsystem. In the control subsystem, a distributed control structurebased on parallel processing of multi-IPC is developed. The timing of rotation,braking, and tracking and snapshotting, the time control strategy of sub-regionalbottle rubbing in the multi-stage is proposed. In addition, the control method of thetracking arm motion is designed. In the image acquisition and processing subsystem,legible images of bottled infusion are acquired after plenty of repeated experimentsfrom the selection of the camera lens to the optical system design. In addition, theworking process and the software architecture of the visual inspection robot systemfor pharmaceutical infusion are analyzed in detail.
3. An FFT frequency domain transform based image calibration and registrationmethod is proposed, testing a series of sequence images to complete the roughalignment of the sequence image parameters. On this basis, through Powelloptimization searching algorithms, the best matching parameters are w orked out.Considering the average cross-relation value of image subset as the cost function ofmatching, the precise image registration parameters are obtained, which achieves thealignment accuracy of the sub-pixel level. In order to reduce the computationalcomplexity of further processing, a method based on probability and statistics for theinterest area extraction of infusion image is proposed. Firstly, the histogramequalization operation is applied to the images to enhance the overall image contrast.Then, the2D maximum entropy threshold segmentation is implemented. Finally,through the method of probability and statistics, the position of the image containingthe liquid region is worked out. In order to suppress the interference of backgroundnoise, an adaptive mean filtering algorithm based on maximum value is proposed.Compared with the traditional filtering algorithms, our algorithm can remove noise successfully while keeping the edge and other details in the images, and relaxing thefuzziness, which can effectively filter noise within the detecting area of the infusionimage.
4. Combined with the characteristics of the inspection for visible foreignsubstances of infusions, an infusion image segmentation algorithm based on CellularNeural Networks (CNN) is proposed. An appropriate CNN template is designed tochange the CNN weighted linear links. Meanwhile, the min/max linking weights ofnonlinear fuzzy operation is introduced, and the structure of Fuzzy Cellular NeuralNetworks (FCNN) is designed. Experiments showed that FCNN was more effectivethan traditional CNN, but FCNN's effect on edge detection was not so perfect. Tosolve this problem, an Improved Fuzzy Cellular Neural Network (IFCNN) is proposedand its convergence and stability are studied. Experimental results showed thatIFCNN can effectively solve the problem appeared in the edge detection that may notbe solved by existing methods, which achieved better approximation to the originalimages.
5. According to the features of complexity and multi-character of the foreignsubstances, a series of morphological, statistical and motion features are selectedfrom the continuous medical image sequences. The calculation method of every usedfeature is introduced. The feature parameters of every target in the experimentalimages are extracted and analyzed. Accordingly, an improved ReliefF algorithm withk nearest neighborhoods is proposed to extract the features. By filtering out irrelevantfeatures with the feature selection algorithm, the algorithm can replace the traditionalnearest neighbor value with the average value of distance to k nearest neighborhoods,to reduce effects of false features such as noises on weight values vastly, and to makefeature extraction more precisely. After analyzing SVM and Boosting algorithms,experiments are carried out to compare the advantages and disadvantages of the twomethods in the classification applications for a variety of foreign substances. Then theAdaBoosting multi-class classification algorithm based on SVM is proposed, andtested by related experiments. In addition, the algorithm is integrated in the softwareof the pharmaceutical defedct inspection line that achieves great outcomes.
6. A sample bottled infusion inspection machine and the related softwareplatform are developed. From the point of engineering application, the softwaresystem of bottled infusion inspection machine is described in detail, ranging fromentrance structure, gripper, rubbing device to electrical control modules, etc. Thedetection and analysis software modules are developed. In the process of the system performance testing, the effectiveness of the bottled infusion inspection machine andrelated algorithms are validated with international verification method (Knapp-Kushner Test). On the other hand, the repeatability, detection accuracy and theperformance on detecting different types of foreign substances of the inspectionmethod are also tested, whose results show that sample machine can meet therequirements of the bottled infusion online inspection effectively.
In this dissertation, through theoretical analysis and experiments, the validityand feasibility of the optical, mechanical, electrical structures and inspectionalgorithms in the proposed bottled infusion visual inspection machine are proved. Asone of the important outcomes of the National High Technology Research andDevelopment Program of China "Intelligent Inspecting and Packaging Robot inLarge-Scale High-Speed Pharmaceutical Automatic Production Line", the develo pedinspection machine is accepted by the experts of the Ministry of Science andTechnology on December9th,2011. The field applications of the inspection machineshow that the developed system can solve most problems in the field applications, andwill play an important role in the automatic detection of pharmaceutical industry, andhave a promising future in a wide range of applications.
引文
[1]胡江宁，张建明．2010年全国医药工业经济运行情况分析．中国医药工业杂志，2011，42（2）：151-156.
[2]郭宏，钟素艳．我国医药包装业的现状及发展对策．医药工程设计，2007，28（1）：51-53.
[3]沈波，冯冬，马爱霞．我国大输液行业存在的问题及对策探讨．药学进展，2008，20（10）：463-467.
[4]崔清新，周婷玉．我国去年人均输液约8瓶，远超国际水平．新华每日电讯，2010-12-25.
[5]曹晓群，张宇，杨慕升．基于质量信息系统的大输液生产中的质量监控．山东理工大学学报(自然科学版)，2010，24(4)：69-71.
[6]郑筱蔓等.中华人民共和国药典.北京:化学工业出版社，2005.
[7]段峰，王耀南，雷晓峰，等.机器视觉技术及其应用综述.自动化博览.2002，19(3):59-61.
[8]贾云得.机器视觉.北京：科学出版社，2000.
[9] Mark Graves, Bruce G. Batchelor.Machine vision for the inspection of naturalproducts,Springer,2003,4-32.
[10] Ramwesh Jain, Rangchar Kasturi, Brain G. Schunck. Machine Vision.McGraw-Hill Companies,Inc,1995.
[11] H. Golnabia, A. Asadpour.Design and application of industrial machine visionsystems.Robotics and Computer-Integrated Manufacturing,2007(23):630–637.
[12] Awcock GJ, Thomas R. Applied image processing. London: Mac Millan NewPress Ltd.;1995.
[13] R. Allen Burns. MACHINE VISION LIGHTING TECHNIQUES FORELECTRONICS ASSEMBLY.http://www.machinevisiononline.org/public/articles/RVSI_LightingforMachineVision.pdf
[14] MVA/SME Lighting and Optics Poster.http://www.sme.org/downloads/mva/mvaposter.pdf
[15] CCD摄像机的选购技巧. http://www.szweiter.com/cn/support/index_1.asp
[16]刘桂玲.大容量注射剂中可见异物的来源及其预防.齐鲁药事,2005,24(12):753-754.
[17]程国义，程念政.用激光散斑场检测药液内杂质含量分布.光子学报.1997，26(2):155-158.
[18] http://www.eisai-mc.co.jp/english/
[19] http://www.seidenader.de/
[20] http://www.brevetti-cea.com/
[21] http://www.gf-industries.it/foodsAndBeverage.php
[22] Akira Ishii, Takayuki Mizuta, and Shigehiko Todo. Detection of foreignsubstances mixed in a plastic bottle of medicinal solution usin g real-time videoimage processing. Pattern Recognition.1998,2:1646-1650.
[23] Herbert Grindinger, Helmut Neusser, Nikolaus Seidenader, et al.Product testingapparatus,US2005/0117149A1.
[24] S. Reed, R. M. Gagliardi, H. M. Shao. Application of Tree-DimensionalFiltering to Moving Target Detection. IEEE Transactions on Aerospace andElectronic Systems.1983,19(6):898-905.
[25] B. A. Porat. Frequency Domain Approach to Multiframe Detection andEstimation of Dim Targets. IEEE Transactions Pattern Analysis and MachineIntelligence,1990,12(4):398-401.
[26] N. C. Mohanty. Computer Tracking of Moving Point Targets in Spaces. IEEETransactions Pattern Analysis and Machine Intelligence,1981,3(5):606-611.
[27] Alexis. P. Tzannes. Detecting Small Moving Object Using Temporal HypothesisTesting. IEEE Transactions on Aerospace and Electronic Systems,2002,38(2):570-586.
[28] Steven D. Blostein, Thomas S. Huang, Detecting Small, Moving Objects inImage Sequences Using Sequential Hypothesis Testing, IEEE Transactions OnSignal Processing,1991,39(7),1611-1629.
[29]张辉，王耀南，周博文.基于机器视觉的液体药品异物检测系统研究.仪器仪表学报.2009，30(3):548-553.
[30]李杨果，王耀南，王威.基于机器视觉的大输液智能灯检机研究.光电工程.2006，33(11):69-74.
[31] M. Carrasco, L. Pizarro, D. Mery. ImageAcquisition and Automated Inspectionof Wine Bottlenecks by Tracking Multiple Views, Proc. Of the8th WSEASInternational Conference on Signal Processing, Computational Geometry andArtificial Vision, Rhodos, Greece,2008,84-89
[32]张辉，王耀南.医药大输液可见异物自动视觉检测方法及系统研究.电子测量与仪器学报，2010，24(2):125-130.
[33]张辉，王耀南.液体药品异物检测智能机器人系统的设计.中国机械工程.2009，20(20):2493-2498.
[34]王耀南，周博文，张辉，等.一种大输液生产线上的自动检测方法及装置.中国发明专利，专利号：ZL200810031626.3.
[35]王耀南，张辉.智能瓶装液体中可见异物在线视觉检测机器人.中国发明专利，申请号：201010120252.X.
[36]刘祥华，张旭.瓶内异物检测设备及其压瓶旋转装置.中国发明专利，申请号：200920176106.1.
[37]刘祥华，张旭.瓶内异物检测设备及其抓瓶旋转装置.中国发明专利，申请号：200910169835.9.
[38]王耀南，马波，张辉，等.用于瓶装液体可见异物视觉检测的搓瓶装置.中国发明专利，申请号：201010193473.X.
[39] Duan Feng, Wang Yaonan. A Machine Vision Inspector for Beer Bottle. Int. J.Engineering Applications of Artificial Intelligence,2007,20(7):1013-1021.
[40] http://www.cis-americas.com/products/machinevision/index.html
[41] http://www.pentax.jp/english/products/catalog/index.html#cctv
[42] http://www.ccs-grp.com/
[43] http://www.cstmv.com/
[44]中华人民共和国国家标准. GB/T2639-2008，玻璃输液瓶.北京：国家技术监督局，2008.
[45]叶斌，彭嘉雄.基于能量累积与顺序形态滤波的红外小目标检测.中国图象图形学报，2002，7(3):251-255.
[46]张兵，卢焕章.序列图像中运动点目标轨迹检测算法研究.电子学报，2004，32(9):1524-1526.
[47] Cheng Y J, Yan H M, Hui H. A novel algorithm for pixel-target detection inliquid image. Acta photonica sinica,2002,31(6):743-747.
[48] Giaime Ginesu,Daniele D. Giusto,Volker M rgner,etal. Detection of ForeignBodies in Food by Thermal Image Processing. IEEE Transactions on IndustrialElectronics,2004,51(2):480-490.
[49] J. Serra, Image Analysis and Mathematical Morphology. New York: Academic,1988.
[50] P.Wallin and P. Haycock, Foreign Body Prevention, Detection and Control.Glasgow, U.K.: Blackie,1998.
[51]章毓晋.图像工程.北京：清华大学出版社，1999.
[52] Yu He, Kim-Hui Yap, Li Chen, etal. A Nonlinear Least Square Technique forSimultaneous Image Registration and Super-Resolution, IEEE Transactions onImage Processing,2007,16(11):2830-2841.
[53] Yang-Ming Zhu, and Steven M. Cochoff.Likelihood Maximization Approach toImage Registration, IEEE Transactions on Image Processing,2002,11(12):1417-1426.
[54] J. B. A. Maintz and M. A. Viergever. A survey of medical image registration,Med. Image Anal,1998,2(1):1-36.
[55] R. Shekhar and V. Zagrodsky. Mutual information-based rigid and nonrigidregistration of ultrasound volumes, IEEE Transactions on Medical Imaging,2002,21:9-22.
[56] Barbara Zitova′, Jan Flusser. Image registration methods: a survey. Image andVision Computing,2003(21):977-1000.
[57] E.De Castro and C.Morandi. Registration of Translated and Rotated Imagesusing Finite Fourier Ti-ansforms. IEEE Transactions Pattern Analysis andMachine Intelligence,1987,(5):700-703.
[58]甘亚莉，涂丹，李国辉.频率域基于梯度预处理的互相关图像配准方法.计算机工程与应用，2007，43（6）：24-26.
[59] Reddy B S,Chatterji B N. An FFT-based technique for translation,rotation,andscale-invariant image registration. IEEE Transactions on Image Processing,1996,5(8):1266-1271.
[60] Manduchi R, Mian G A. Accuracy analysis for correlation-based imageregistration algorithms.Proc ISCAS,1993:834-837.
[61] Min Xu,Pramod K. Varshney.A Subspace Method for Fourier-Based ImageRegistration. IEEE Geoscience and Remote Sensing Letters,2009,6(3):491-494.
[62] Georgios Tzimiropoulos,Vasileios Argyriou,Stefanos Zafeiriou, etal. RobustFFT-Based Scale-Invariant Image Registration with Image Gradients,IEEETransactions on Pattern Analysis and Machine Intelligence,2010,32(10):1899-1906.
[63]温江涛，王伯雄，秦垚.基于局部灰度梯度特征的图像快速配准方法.清华大学学报（自然科学版），2009，49(5):673-675.
[64] Manuel Guizar-Sicairos, Samuel T. Thurman, and James R. Fienup,Efficientsubpixel image registration algorithms.OPTICS LETTERS,2008,33(2):156-158.
[65]张广军.视觉测量.北京:科学出版社，2008.
[66] DING Hai-yong, BIAN Zheng-fu. A Sub-pixel Registration Approach Based onPowell Algorithm. Acta Photonica Sinica,2009,38(12):3322-3327.
[67]李弼程，彭天强，彭波，等．智能图像处理技术．北京：电子工业出版社．2004.
[68]张光澄，黄世莹，侯泽华.最优化计算方法.成都：成都科技大学出版社，1989，47-140.
[69]游兆永，徐成贤，吴振国，等.实用最优化方法.天津：天津科技翻译出版公司，1990，45-127.
[70]赵海峰，姚丽莎，罗斌.改进的人工鱼群算法和Powell法结合的医学图像配准.西安交通大学学报，2011，45(4):46-52.
[71]肖开明.图像配准算法及其在印刷质量检测中的应用：[上海大学硕士学位论文].上海：上海大学，2004，45-66.
[72] XU Xiaoyan, DONY R D. Differential evolution with Powell's direction setmethod in medical image regeistration.Proceedings of the IEEE InternationalSymposium on Biomedical Imaging:Macro to Nano.Arlington,USA,2004.
[73]夏召强，冯晓毅，彭进业.基于边缘与深度特征的感兴趣区域检测技术.计算机仿真，2009，26（7）：248-251.
[74] V N avalpakkam and L Itt. An integrated model of top-down and bottom-upattention for optimizing detection speed. CVPR,2006.2049-2056.
[75]张鹏，王润生.静态图像中的感兴趣区域检测技术.中国图像图形学报，2005，10(2)：142-148.
[76] Tao Wen-Bing; Tian Jin-Wen, Liu Jian. Image segmentation by three-levelthresholding based on maximum fuzzy entropy and genetic algorithm. PatternRecognition Letters,2003,24(16):3069-3078.
[77] A.D. Brink. Thresholding of digital images using two-dimensional entropies.Pattern Recognition,1992,(25):803-808.
[78] M. Oral and U. Deniz. Centre of mass model–A novel approach to backgroundmodeling for segmentation of moving objects. Image Vision Computing.2007,25(8):1365-1376.
[79]张飞，李承芳，史丽娜，等.复杂背景下运动点目标的检测算法.光学技术,2005，31(1):55-61.
[80] Wang Zhou， Zhang David． Progressive Switching Median Filter for theRemoval of Impulse Noise from Highly Corrupted Images. IEEE Transactionson Circuits and Systems Part II: Analog and Digital Signal Processing,1999,46(1):23-25.
[81] Wang Jung-Hua, Lin Lian-Da. Improved Median Filter Using Min-maxAlgorithm for Image Processing. Electronics Letters,1997,33(16):211-215.
[82]叶斌，彭嘉雄.基于能量累积与顺序形态滤波的红外小目标检测.中国图象图形学报,2002,7(3):251-255.
[83]杨樊，韩艳丽.一种基于极值的自适应均值滤波算法.红外与激光工程,2006(35):116-120.
[84] Nikos Paragios, Yunmei Chen,Olivier Faugeras. Handbook of MathematicalModels in Computer Vision. Springer,2006.
[85] Chua L O, YANG L. Cellular neural networks:Theory. IEEE Transactions onCircuits and Systems,1988,35(10):1257-1272.
[86] Chua L O, Yang L. Cellular neural networks:Applications. IEEE Transactionson Circuits and Systems,1988,32(10):1273-1290.
[87] Chua L O. Cellular neural networks: a vision of complexity.InternationalJournal Bifurcation and Chaos,1997,7(10):2219-2425.
[88]刘万军，姜庆玲，张闯.基于CNN彩色图像边缘检测的车牌定位方法.自动化学报，2009，35（12）：1503-1512.
[89]冯强，于盛林，王怀颖，等.基于细胞神经网络的有核细胞边缘检测方法研究.中国生物医学工程学报，2008，27(1):29-32.
[90]徐国保，洪丽兰，郝彦爽，等.一种用细胞神经网提取遥感图像边缘的新方法.计算机应用研究，2008，25(11):3504-3506.
[91]姚力，刘佳敏，谢咏圭，等.基于细胞神经网络的图像分割及其在医学图像中的应用.中国科学(E辑)，2001，31:167-171.
[92] D. Rodr′iguez-Fern′andez, D.L.Vilari no and X.M.Pardo.CNN Implementationof a Moving Object Segmentation Approach for Real-Time Video Surveillance.11th International Workshop on Cellular Neural Networks and theirApplications,2008,129-134.
[93] Guido Vagliasindi, Andrea Murari, Paolo Arena, and etal. Cellular NeuralNetwork Algorithms for Real-Time Image Analysis in Plasma Fusion. IEEETransactions on Instrumentation and Measurement,2009,58(8):2417-2425.
[94] K. R. Crounse and L. O. Chua. Methods for Image Processing in CellularNeural Networks: A Tutorial. IEEE Transactions on Circuits and Systems,1995,42(10):583-601.
[95] I. Szatmari, A. Schultz,Cs. Rekeczky,etal. Morphology and Autowave Metric onCNN Applied to Bubble-Debris Classification. IEEE Transactions on NeuralNetworks,2000,11(6):1385-1393.
[96] Cs. Rekeczky, A. Schultz, I. Szatmari,etal.Image Segmentation and EdgeDetection via Constrained Diffusion and Adaptive Morphology: a CNNApproach to Bubble/Debris Image Enhancement, Memorandum UCB/ERLM97/96, Electronic Research Laboratory, University of California at Berkeley,December1997.
[97] I. Szatmari, A. Schultz, Cs. Rekeczky,etal.Bubble-Debris Classification viaBinary Morphology and Autowave Metric on CNN, Berkeley Memo,Electronics Research Laboratory, University of California, Berkeley, M97/97,1997.
[98] Wener BorgesSampaio, EdgarMoraesDiniz, Aristo′fanes Correa Silva andetal.Detection of masses in mammogram images using CNN,geostatisticfunctions and SVM. Computers in Biology and Medicine,2011,(41):653-664.
[99] Yang T, Yang LB. Fuzzy Cellular Neural Network:Theory.Proc Int'l Workshopon Cellular Neural Networks and Their Applications. NewYork: IEEE,1996:225-230.
[100] Tao Y, Yang L B. The Global Stability of Fuzzy Cellular Neural Network. IEEETransactions on Circuits and Systems,1996,(43):880-883.
[101] Yuan Yao,et al, Fuzzy Cellular Neural Network and Its Application in ChineseCharacter Reconstruction. Chinese J. Computer Research and Development,1999,36(3):282-286.
[102] T. Yang, L. B. Yang, C. W. Wu, and L. O. Chua, Fuzzy cellular neural networks:Applications, in Proc. Cellular Neural Networks Application (CNNA’96),225–230.
[103] T.Yang.Handbook of CNN Image Processing:All you need to know aboutcellular neural networks.Yang's Scientific Press,2003,190-197(http://www.yangsky.com).
[104] Chin-Teng Lin, Chun-Lung Chang, Wen-Chang Cheng. A Recurrent FuzzyCellular Neural Network System With Automatic Structure and TemplateLearning. IEEE Transactions on Circuits and Systems I: Regular Papers,2004,51(5):1024-1035.
[105]付端，王士同，胡德文.改进的模糊细胞神经网络(IFCNN)的应用与研究.控制与决策，2006，21(1):114-117.
[106] Wang Shitong, Korris F.L.Chung, Fu Duan.Applying the improved fuzzycellular neural network IFCNN to white blood cell detection. Neurocomputing,2007,(70):1348–1359.
[107] Wang Shitong, Fu Duan, Xu Min,and etal. Advanced fuzzy cellular neuralnetwork: Application to CT liver images. Artificial Intelligence in Medicine,2007,(39),65-77.
[108]陈安平，袁顺莲.模糊细胞神经网络的全局稳定性.郴州师范高等专科学校学报，2001，22(5):6-9.
[109] Tao Yang, Linbao Yang. The Global Stability of Fuzzy Cellular Neura l Network.IEEE Transactions on Circuits and Systems I: Fundamental Theory andApplications,1996,(43):880-883.
[110]杨福刚.输液中微小异物目标视觉检测技术研究.[山东大学博士学位论文].济南：山东大学，2008.
[111]成玉娟.液体中小目标检测算法研究及应用.[浙江大学硕士学位论文].杭州：浙江大学，2002.
[112]王大千.基于机器视觉的医药液体制剂异物检测算法研究.[南京大学硕士学位论文].南京：南京大学，2011.
[113]李伟.基于机器视觉的安瓿内可见异物检测系统的研究.[清华大学硕士学位论文].北京：清华大学，2010.
[114]孙即祥等.模式识别中的特征提取与计算机视觉不变量.北京：国防工业出版社，2001.
[115]王润生.图像理解.北京:国防科技大学出版社，1995.
[116] Cortes, C., Vapnik,V. Support-vector networks. Machine Learning，1995，20(3),273-297.
[117] Kim H C, Pang S, Je HM, et al. Constructing Support Vector Machine Ensemble.Pattern Recognition,2003,36(12):2757-2767.
[118] Valentini G, Dietterich T G. Bias-variance Analysis of Support Vector Machinesfor the Development of SVM-Based Ensemble Methods. Journal of MachineLearning Research,2004,5:725-775.
[119] R. E. Schapire. A brief introduction to boosting. In: Proceedings of the16thInternational Joint Conference on Artificial Intelligence. Stockholm, Sweden:Morgan Kaufmann Publishers,1999,14011406.
[120] R. E. Schapire, Y. Freund, P. Bartlett. Boosting the margin: a new explanationfor the effectiveness of voting methods. Ann. Stat.,1998,26(5):1651–1686.
[121] M.K.Hu.Visual Pattern Recognition by Moment Invariants. IEEE Transactionson Information Theory,1962,(8):179-187.
[122] Wong Y R.Scene Matching with Invariant Moments.Computer Graphics andImage Processing,1978：16-24.
[123] R.C.Gonzalez,R.E.Woods. Digital Image Processing (英文版).北京：电子工业出版社.2002.7：672-675.
[124]陆方杰，夏顺仁.基于归一化转动惯量的显微图像拼接算法.中国医疗器械杂志，2007，31(6):404-406.
[125] K.Kira,L. A.Rendell.A Practical Approach to Feature Selection. Proceedings ofthe9th International Workshop on Machine Learning, San Francisco,1992:249-256.
[126] I.Kononenko. Estimating Attributes Analysis and Extensions of Relief.Proceedings of the7th European Conference on Machine Learni ng, Berlin:Springer,1994:171-182.
[127] L.G.Valiant.A Theory of the Learnable.Communication of ACM.1984,27(11):1134-1142.
[128] R.E.Schapire. The Strength of Weak Leaning Ability. Machine Learning.1990,5(2):197-227.
[129] Y.Freund. Boosting a Weak Learning Algorithm by Majority. Information andComputation.1995,121(2):256-285.
[130] Drucker H, Schapire R, Simard P. Boosting performance in neural networks.International Journal of Pattern Recognition and Artificial Intelligence,1993,7(4):705-719.
[131] Freund, Y., Schapire, R.E. A decision-theoretic generalization of on-linelearning and an application to boosting. Journal of Computer and SystemSciences,1997,55(1),119-139.
[132] Freund, Y., Schapire, R.E. Improved boosting algorithms using confidenceratedpredictions. Machine Learning.1999,37(3),297-336.
[133] Robert A. Ochs, Jonathan G. Goldin, Fereidoun Abtin and etal. Automatedclassification of lung bronchovascular anatomy in CT using AdaBoost. MedicalImage Analysis,2007(11):315-324.
[134] Yijun Sun, Sinisa Todorovic, Jian Li. Unifying multi-class AdaBoost algorithmswith binary base learners under the margin framework. Pattern RecognitionLetters,2007(28):631-643.
[135]沈志熙，黄席樾，杨镇宇，等.基于Boosting的智能车辆多类障碍物识别.计算机工程，2009，35（14）：241-243.
[136] GE Jun-Feng, LUO Yu-Pin. A Comprehensive Study for Asymmetric AdaBoostand Its Application in Object Detection. Acta Automatica Sinica,2009,35(11):1403-1409.
[137] S.K. Mathanker, P.R. Weckler. AdaBoost classifiers for pecan defectclassification. Computers and Electronics in Agriculture,2011,(77):60-68.
[138] Vezhnevets, A.(2009). GML AdaBoost Matlab Toolbox0.3..
[139] Hatice Dogan, Olcay Akay. Using AdaBoost classifiers in a hierarchicalframework for classifying surface images of marble slabs. Expert Systems withApplications,2010(37)8814-8821.
[140] H. Yin, Y. Cao, H. Sun. Combining pyramid representation and AdaBoost forurban scene classification using high-resolution synthetic aperture radar images.IET Radar, Sonar and Navigation,2011,5(1):58-64.
[141] Hearst M A，Dumais S T，OsroanE，etal. Support Vetor Machines. IEEEIntelligent Systems,1998,13(4):18-28.
[142] Hsu C,Lin C.A Comparison of Methods for Multiclass Support Vector Machines.IEEE Transactions on Neural Networks,2002,13(2):415-425.
[143] Lin Chun-Fu, Wang Sheng-De. Fuzzy support vector machines. IEEETransactions on Neural Networks,2002,13(2):464-471
[144] O. Chapelle, P. Haffner, and V. N. Vapnik. Support vector machines forhistogram-based image classification. IEEE Transactions on Neural Networks,1999,10(5):1055-1064.
[145] J.J. Rodriguez， J. Maudes. Boosting Reeombined Weak Classifiers. PatterRecognition Letters,2008,29:1049-1059.
[146] Mattia Marconcini, Gustavo Camps-Valls, and Lorenzo Bruzzone. A CompositeSemisupervised SVM for Classification of Hyperspectral Images. IEEE Trans.Geosci. Remote Sens.2009,6(2):234-238.
[147] Asuncion,A.&Newman,D.J.(2007).UCI Machine Learning Repository
[www.ics.uci.edu/~mlearn/MLRepository.html].Irvine,CA:University ofCalifornia, School of Information and Computer Science.
[148] J.J. Rodriguez， J. Maudes. Boosting Recombined Weak Classifiers. PatterRecognition Letters,2008,29:1049-1059.
[149] Xuchun Li, Lei Wang. AdaBoost with SVM-based component classifiers.Engineering Applications of Artificial Intelligence,2008(21):785–795.
[150] E. Romero, L. Marquez, X. Carreras, Margin maximization with feed-forwardneural networks: a comparative study with SVM and AdaBoost. Neuro-computing,2004,(57):313-344.
[151]何斌，马大予，王运坚，等.visual c++数字图像处理(第二版).北京:人民邮电出版社出版社，2002.
[2]郭宏，钟素艳．我国医药包装业的现状及发展对策．医药工程设计，2007，28（1）：51-53.
[3]沈波，冯冬，马爱霞．我国大输液行业存在的问题及对策探讨．药学进展，2008，20（10）：463-467.
[4]崔清新，周婷玉．我国去年人均输液约8瓶，远超国际水平．新华每日电讯，2010-12-25.
[5]曹晓群，张宇，杨慕升．基于质量信息系统的大输液生产中的质量监控．山东理工大学学报(自然科学版)，2010，24(4)：69-71.
[6]郑筱蔓等.中华人民共和国药典.北京:化学工业出版社，2005.
[7]段峰，王耀南，雷晓峰，等.机器视觉技术及其应用综述.自动化博览.2002，19(3):59-61.
[8]贾云得.机器视觉.北京：科学出版社，2000.
[9] Mark Graves, Bruce G. Batchelor.Machine vision for the inspection of naturalproducts,Springer,2003,4-32.
[10] Ramwesh Jain, Rangchar Kasturi, Brain G. Schunck. Machine Vision.McGraw-Hill Companies,Inc,1995.
[11] H. Golnabia, A. Asadpour.Design and application of industrial machine visionsystems.Robotics and Computer-Integrated Manufacturing,2007(23):630–637.
[12] Awcock GJ, Thomas R. Applied image processing. London: Mac Millan NewPress Ltd.;1995.
[13] R. Allen Burns. MACHINE VISION LIGHTING TECHNIQUES FORELECTRONICS ASSEMBLY.http://www.machinevisiononline.org/public/articles/RVSI_LightingforMachineVision.pdf
[14] MVA/SME Lighting and Optics Poster.http://www.sme.org/downloads/mva/mvaposter.pdf
[15] CCD摄像机的选购技巧. http://www.szweiter.com/cn/support/index_1.asp
[16]刘桂玲.大容量注射剂中可见异物的来源及其预防.齐鲁药事,2005,24(12):753-754.
[17]程国义，程念政.用激光散斑场检测药液内杂质含量分布.光子学报.1997，26(2):155-158.
[18] http://www.eisai-mc.co.jp/english/
[19] http://www.seidenader.de/
[20] http://www.brevetti-cea.com/
[21] http://www.gf-industries.it/foodsAndBeverage.php
[22] Akira Ishii, Takayuki Mizuta, and Shigehiko Todo. Detection of foreignsubstances mixed in a plastic bottle of medicinal solution usin g real-time videoimage processing. Pattern Recognition.1998,2:1646-1650.
[23] Herbert Grindinger, Helmut Neusser, Nikolaus Seidenader, et al.Product testingapparatus,US2005/0117149A1.
[24] S. Reed, R. M. Gagliardi, H. M. Shao. Application of Tree-DimensionalFiltering to Moving Target Detection. IEEE Transactions on Aerospace andElectronic Systems.1983,19(6):898-905.
[25] B. A. Porat. Frequency Domain Approach to Multiframe Detection andEstimation of Dim Targets. IEEE Transactions Pattern Analysis and MachineIntelligence,1990,12(4):398-401.
[26] N. C. Mohanty. Computer Tracking of Moving Point Targets in Spaces. IEEETransactions Pattern Analysis and Machine Intelligence,1981,3(5):606-611.
[27] Alexis. P. Tzannes. Detecting Small Moving Object Using Temporal HypothesisTesting. IEEE Transactions on Aerospace and Electronic Systems,2002,38(2):570-586.
[28] Steven D. Blostein, Thomas S. Huang, Detecting Small, Moving Objects inImage Sequences Using Sequential Hypothesis Testing, IEEE Transactions OnSignal Processing,1991,39(7),1611-1629.
[29]张辉，王耀南，周博文.基于机器视觉的液体药品异物检测系统研究.仪器仪表学报.2009，30(3):548-553.
[30]李杨果，王耀南，王威.基于机器视觉的大输液智能灯检机研究.光电工程.2006，33(11):69-74.
[31] M. Carrasco, L. Pizarro, D. Mery. ImageAcquisition and Automated Inspectionof Wine Bottlenecks by Tracking Multiple Views, Proc. Of the8th WSEASInternational Conference on Signal Processing, Computational Geometry andArtificial Vision, Rhodos, Greece,2008,84-89
[32]张辉，王耀南.医药大输液可见异物自动视觉检测方法及系统研究.电子测量与仪器学报，2010，24(2):125-130.
[33]张辉，王耀南.液体药品异物检测智能机器人系统的设计.中国机械工程.2009，20(20):2493-2498.
[34]王耀南，周博文，张辉，等.一种大输液生产线上的自动检测方法及装置.中国发明专利，专利号：ZL200810031626.3.
[35]王耀南，张辉.智能瓶装液体中可见异物在线视觉检测机器人.中国发明专利，申请号：201010120252.X.
[36]刘祥华，张旭.瓶内异物检测设备及其压瓶旋转装置.中国发明专利，申请号：200920176106.1.
[37]刘祥华，张旭.瓶内异物检测设备及其抓瓶旋转装置.中国发明专利，申请号：200910169835.9.
[38]王耀南，马波，张辉，等.用于瓶装液体可见异物视觉检测的搓瓶装置.中国发明专利，申请号：201010193473.X.
[39] Duan Feng, Wang Yaonan. A Machine Vision Inspector for Beer Bottle. Int. J.Engineering Applications of Artificial Intelligence,2007,20(7):1013-1021.
[40] http://www.cis-americas.com/products/machinevision/index.html
[41] http://www.pentax.jp/english/products/catalog/index.html#cctv
[42] http://www.ccs-grp.com/
[43] http://www.cstmv.com/
[44]中华人民共和国国家标准. GB/T2639-2008，玻璃输液瓶.北京：国家技术监督局，2008.
[45]叶斌，彭嘉雄.基于能量累积与顺序形态滤波的红外小目标检测.中国图象图形学报，2002，7(3):251-255.
[46]张兵，卢焕章.序列图像中运动点目标轨迹检测算法研究.电子学报，2004，32(9):1524-1526.
[47] Cheng Y J, Yan H M, Hui H. A novel algorithm for pixel-target detection inliquid image. Acta photonica sinica,2002,31(6):743-747.
[48] Giaime Ginesu,Daniele D. Giusto,Volker M rgner,etal. Detection of ForeignBodies in Food by Thermal Image Processing. IEEE Transactions on IndustrialElectronics,2004,51(2):480-490.
[49] J. Serra, Image Analysis and Mathematical Morphology. New York: Academic,1988.
[50] P.Wallin and P. Haycock, Foreign Body Prevention, Detection and Control.Glasgow, U.K.: Blackie,1998.
[51]章毓晋.图像工程.北京：清华大学出版社，1999.
[52] Yu He, Kim-Hui Yap, Li Chen, etal. A Nonlinear Least Square Technique forSimultaneous Image Registration and Super-Resolution, IEEE Transactions onImage Processing,2007,16(11):2830-2841.
[53] Yang-Ming Zhu, and Steven M. Cochoff.Likelihood Maximization Approach toImage Registration, IEEE Transactions on Image Processing,2002,11(12):1417-1426.
[54] J. B. A. Maintz and M. A. Viergever. A survey of medical image registration,Med. Image Anal,1998,2(1):1-36.
[55] R. Shekhar and V. Zagrodsky. Mutual information-based rigid and nonrigidregistration of ultrasound volumes, IEEE Transactions on Medical Imaging,2002,21:9-22.
[56] Barbara Zitova′, Jan Flusser. Image registration methods: a survey. Image andVision Computing,2003(21):977-1000.
[57] E.De Castro and C.Morandi. Registration of Translated and Rotated Imagesusing Finite Fourier Ti-ansforms. IEEE Transactions Pattern Analysis andMachine Intelligence,1987,(5):700-703.
[58]甘亚莉，涂丹，李国辉.频率域基于梯度预处理的互相关图像配准方法.计算机工程与应用，2007，43（6）：24-26.
[59] Reddy B S,Chatterji B N. An FFT-based technique for translation,rotation,andscale-invariant image registration. IEEE Transactions on Image Processing,1996,5(8):1266-1271.
[60] Manduchi R, Mian G A. Accuracy analysis for correlation-based imageregistration algorithms.Proc ISCAS,1993:834-837.
[61] Min Xu,Pramod K. Varshney.A Subspace Method for Fourier-Based ImageRegistration. IEEE Geoscience and Remote Sensing Letters,2009,6(3):491-494.
[62] Georgios Tzimiropoulos,Vasileios Argyriou,Stefanos Zafeiriou, etal. RobustFFT-Based Scale-Invariant Image Registration with Image Gradients,IEEETransactions on Pattern Analysis and Machine Intelligence,2010,32(10):1899-1906.
[63]温江涛，王伯雄，秦垚.基于局部灰度梯度特征的图像快速配准方法.清华大学学报（自然科学版），2009，49(5):673-675.
[64] Manuel Guizar-Sicairos, Samuel T. Thurman, and James R. Fienup,Efficientsubpixel image registration algorithms.OPTICS LETTERS,2008,33(2):156-158.
[65]张广军.视觉测量.北京:科学出版社，2008.
[66] DING Hai-yong, BIAN Zheng-fu. A Sub-pixel Registration Approach Based onPowell Algorithm. Acta Photonica Sinica,2009,38(12):3322-3327.
[67]李弼程，彭天强，彭波，等．智能图像处理技术．北京：电子工业出版社．2004.
[68]张光澄，黄世莹，侯泽华.最优化计算方法.成都：成都科技大学出版社，1989，47-140.
[69]游兆永，徐成贤，吴振国，等.实用最优化方法.天津：天津科技翻译出版公司，1990，45-127.
[70]赵海峰，姚丽莎，罗斌.改进的人工鱼群算法和Powell法结合的医学图像配准.西安交通大学学报，2011，45(4):46-52.
[71]肖开明.图像配准算法及其在印刷质量检测中的应用：[上海大学硕士学位论文].上海：上海大学，2004，45-66.
[72] XU Xiaoyan, DONY R D. Differential evolution with Powell's direction setmethod in medical image regeistration.Proceedings of the IEEE InternationalSymposium on Biomedical Imaging:Macro to Nano.Arlington,USA,2004.
[73]夏召强，冯晓毅，彭进业.基于边缘与深度特征的感兴趣区域检测技术.计算机仿真，2009，26（7）：248-251.
[74] V N avalpakkam and L Itt. An integrated model of top-down and bottom-upattention for optimizing detection speed. CVPR,2006.2049-2056.
[75]张鹏，王润生.静态图像中的感兴趣区域检测技术.中国图像图形学报，2005，10(2)：142-148.
[76] Tao Wen-Bing; Tian Jin-Wen, Liu Jian. Image segmentation by three-levelthresholding based on maximum fuzzy entropy and genetic algorithm. PatternRecognition Letters,2003,24(16):3069-3078.
[77] A.D. Brink. Thresholding of digital images using two-dimensional entropies.Pattern Recognition,1992,(25):803-808.
[78] M. Oral and U. Deniz. Centre of mass model–A novel approach to backgroundmodeling for segmentation of moving objects. Image Vision Computing.2007,25(8):1365-1376.
[79]张飞，李承芳，史丽娜，等.复杂背景下运动点目标的检测算法.光学技术,2005，31(1):55-61.
[80] Wang Zhou， Zhang David． Progressive Switching Median Filter for theRemoval of Impulse Noise from Highly Corrupted Images. IEEE Transactionson Circuits and Systems Part II: Analog and Digital Signal Processing,1999,46(1):23-25.
[81] Wang Jung-Hua, Lin Lian-Da. Improved Median Filter Using Min-maxAlgorithm for Image Processing. Electronics Letters,1997,33(16):211-215.
[82]叶斌，彭嘉雄.基于能量累积与顺序形态滤波的红外小目标检测.中国图象图形学报,2002,7(3):251-255.
[83]杨樊，韩艳丽.一种基于极值的自适应均值滤波算法.红外与激光工程,2006(35):116-120.
[84] Nikos Paragios, Yunmei Chen,Olivier Faugeras. Handbook of MathematicalModels in Computer Vision. Springer,2006.
[85] Chua L O, YANG L. Cellular neural networks:Theory. IEEE Transactions onCircuits and Systems,1988,35(10):1257-1272.
[86] Chua L O, Yang L. Cellular neural networks:Applications. IEEE Transactionson Circuits and Systems,1988,32(10):1273-1290.
[87] Chua L O. Cellular neural networks: a vision of complexity.InternationalJournal Bifurcation and Chaos,1997,7(10):2219-2425.
[88]刘万军，姜庆玲，张闯.基于CNN彩色图像边缘检测的车牌定位方法.自动化学报，2009，35（12）：1503-1512.
[89]冯强，于盛林，王怀颖，等.基于细胞神经网络的有核细胞边缘检测方法研究.中国生物医学工程学报，2008，27(1):29-32.
[90]徐国保，洪丽兰，郝彦爽，等.一种用细胞神经网提取遥感图像边缘的新方法.计算机应用研究，2008，25(11):3504-3506.
[91]姚力，刘佳敏，谢咏圭，等.基于细胞神经网络的图像分割及其在医学图像中的应用.中国科学(E辑)，2001，31:167-171.
[92] D. Rodr′iguez-Fern′andez, D.L.Vilari no and X.M.Pardo.CNN Implementationof a Moving Object Segmentation Approach for Real-Time Video Surveillance.11th International Workshop on Cellular Neural Networks and theirApplications,2008,129-134.
[93] Guido Vagliasindi, Andrea Murari, Paolo Arena, and etal. Cellular NeuralNetwork Algorithms for Real-Time Image Analysis in Plasma Fusion. IEEETransactions on Instrumentation and Measurement,2009,58(8):2417-2425.
[94] K. R. Crounse and L. O. Chua. Methods for Image Processing in CellularNeural Networks: A Tutorial. IEEE Transactions on Circuits and Systems,1995,42(10):583-601.
[95] I. Szatmari, A. Schultz,Cs. Rekeczky,etal. Morphology and Autowave Metric onCNN Applied to Bubble-Debris Classification. IEEE Transactions on NeuralNetworks,2000,11(6):1385-1393.
[96] Cs. Rekeczky, A. Schultz, I. Szatmari,etal.Image Segmentation and EdgeDetection via Constrained Diffusion and Adaptive Morphology: a CNNApproach to Bubble/Debris Image Enhancement, Memorandum UCB/ERLM97/96, Electronic Research Laboratory, University of California at Berkeley,December1997.
[97] I. Szatmari, A. Schultz, Cs. Rekeczky,etal.Bubble-Debris Classification viaBinary Morphology and Autowave Metric on CNN, Berkeley Memo,Electronics Research Laboratory, University of California, Berkeley, M97/97,1997.
[98] Wener BorgesSampaio, EdgarMoraesDiniz, Aristo′fanes Correa Silva andetal.Detection of masses in mammogram images using CNN,geostatisticfunctions and SVM. Computers in Biology and Medicine,2011,(41):653-664.
[99] Yang T, Yang LB. Fuzzy Cellular Neural Network:Theory.Proc Int'l Workshopon Cellular Neural Networks and Their Applications. NewYork: IEEE,1996:225-230.
[100] Tao Y, Yang L B. The Global Stability of Fuzzy Cellular Neural Network. IEEETransactions on Circuits and Systems,1996,(43):880-883.
[101] Yuan Yao,et al, Fuzzy Cellular Neural Network and Its Application in ChineseCharacter Reconstruction. Chinese J. Computer Research and Development,1999,36(3):282-286.
[102] T. Yang, L. B. Yang, C. W. Wu, and L. O. Chua, Fuzzy cellular neural networks:Applications, in Proc. Cellular Neural Networks Application (CNNA’96),225–230.
[103] T.Yang.Handbook of CNN Image Processing:All you need to know aboutcellular neural networks.Yang's Scientific Press,2003,190-197(http://www.yangsky.com).
[104] Chin-Teng Lin, Chun-Lung Chang, Wen-Chang Cheng. A Recurrent FuzzyCellular Neural Network System With Automatic Structure and TemplateLearning. IEEE Transactions on Circuits and Systems I: Regular Papers,2004,51(5):1024-1035.
[105]付端，王士同，胡德文.改进的模糊细胞神经网络(IFCNN)的应用与研究.控制与决策，2006，21(1):114-117.
[106] Wang Shitong, Korris F.L.Chung, Fu Duan.Applying the improved fuzzycellular neural network IFCNN to white blood cell detection. Neurocomputing,2007,(70):1348–1359.
[107] Wang Shitong, Fu Duan, Xu Min,and etal. Advanced fuzzy cellular neuralnetwork: Application to CT liver images. Artificial Intelligence in Medicine,2007,(39),65-77.
[108]陈安平，袁顺莲.模糊细胞神经网络的全局稳定性.郴州师范高等专科学校学报，2001，22(5):6-9.
[109] Tao Yang, Linbao Yang. The Global Stability of Fuzzy Cellular Neura l Network.IEEE Transactions on Circuits and Systems I: Fundamental Theory andApplications,1996,(43):880-883.
[110]杨福刚.输液中微小异物目标视觉检测技术研究.[山东大学博士学位论文].济南：山东大学，2008.
[111]成玉娟.液体中小目标检测算法研究及应用.[浙江大学硕士学位论文].杭州：浙江大学，2002.
[112]王大千.基于机器视觉的医药液体制剂异物检测算法研究.[南京大学硕士学位论文].南京：南京大学，2011.
[113]李伟.基于机器视觉的安瓿内可见异物检测系统的研究.[清华大学硕士学位论文].北京：清华大学，2010.
[114]孙即祥等.模式识别中的特征提取与计算机视觉不变量.北京：国防工业出版社，2001.
[115]王润生.图像理解.北京:国防科技大学出版社，1995.
[116] Cortes, C., Vapnik,V. Support-vector networks. Machine Learning，1995，20(3),273-297.
[117] Kim H C, Pang S, Je HM, et al. Constructing Support Vector Machine Ensemble.Pattern Recognition,2003,36(12):2757-2767.
[118] Valentini G, Dietterich T G. Bias-variance Analysis of Support Vector Machinesfor the Development of SVM-Based Ensemble Methods. Journal of MachineLearning Research,2004,5:725-775.
[119] R. E. Schapire. A brief introduction to boosting. In: Proceedings of the16thInternational Joint Conference on Artificial Intelligence. Stockholm, Sweden:Morgan Kaufmann Publishers,1999,14011406.
[120] R. E. Schapire, Y. Freund, P. Bartlett. Boosting the margin: a new explanationfor the effectiveness of voting methods. Ann. Stat.,1998,26(5):1651–1686.
[121] M.K.Hu.Visual Pattern Recognition by Moment Invariants. IEEE Transactionson Information Theory,1962,(8):179-187.
[122] Wong Y R.Scene Matching with Invariant Moments.Computer Graphics andImage Processing,1978：16-24.
[123] R.C.Gonzalez,R.E.Woods. Digital Image Processing (英文版).北京：电子工业出版社.2002.7：672-675.
[124]陆方杰，夏顺仁.基于归一化转动惯量的显微图像拼接算法.中国医疗器械杂志，2007，31(6):404-406.
[125] K.Kira,L. A.Rendell.A Practical Approach to Feature Selection. Proceedings ofthe9th International Workshop on Machine Learning, San Francisco,1992:249-256.
[126] I.Kononenko. Estimating Attributes Analysis and Extensions of Relief.Proceedings of the7th European Conference on Machine Learni ng, Berlin:Springer,1994:171-182.
[127] L.G.Valiant.A Theory of the Learnable.Communication of ACM.1984,27(11):1134-1142.
[128] R.E.Schapire. The Strength of Weak Leaning Ability. Machine Learning.1990,5(2):197-227.
[129] Y.Freund. Boosting a Weak Learning Algorithm by Majority. Information andComputation.1995,121(2):256-285.
[130] Drucker H, Schapire R, Simard P. Boosting performance in neural networks.International Journal of Pattern Recognition and Artificial Intelligence,1993,7(4):705-719.
[131] Freund, Y., Schapire, R.E. A decision-theoretic generalization of on-linelearning and an application to boosting. Journal of Computer and SystemSciences,1997,55(1),119-139.
[132] Freund, Y., Schapire, R.E. Improved boosting algorithms using confidenceratedpredictions. Machine Learning.1999,37(3),297-336.
[133] Robert A. Ochs, Jonathan G. Goldin, Fereidoun Abtin and etal. Automatedclassification of lung bronchovascular anatomy in CT using AdaBoost. MedicalImage Analysis,2007(11):315-324.
[134] Yijun Sun, Sinisa Todorovic, Jian Li. Unifying multi-class AdaBoost algorithmswith binary base learners under the margin framework. Pattern RecognitionLetters,2007(28):631-643.
[135]沈志熙，黄席樾，杨镇宇，等.基于Boosting的智能车辆多类障碍物识别.计算机工程，2009，35（14）：241-243.
[136] GE Jun-Feng, LUO Yu-Pin. A Comprehensive Study for Asymmetric AdaBoostand Its Application in Object Detection. Acta Automatica Sinica,2009,35(11):1403-1409.
[137] S.K. Mathanker, P.R. Weckler. AdaBoost classifiers for pecan defectclassification. Computers and Electronics in Agriculture,2011,(77):60-68.
[138] Vezhnevets, A.(2009). GML AdaBoost Matlab Toolbox0.3.
[139] Hatice Dogan, Olcay Akay. Using AdaBoost classifiers in a hierarchicalframework for classifying surface images of marble slabs. Expert Systems withApplications,2010(37)8814-8821.
[140] H. Yin, Y. Cao, H. Sun. Combining pyramid representation and AdaBoost forurban scene classification using high-resolution synthetic aperture radar images.IET Radar, Sonar and Navigation,2011,5(1):58-64.
[141] Hearst M A，Dumais S T，OsroanE，etal. Support Vetor Machines. IEEEIntelligent Systems,1998,13(4):18-28.
[142] Hsu C,Lin C.A Comparison of Methods for Multiclass Support Vector Machines.IEEE Transactions on Neural Networks,2002,13(2):415-425.
[143] Lin Chun-Fu, Wang Sheng-De. Fuzzy support vector machines. IEEETransactions on Neural Networks,2002,13(2):464-471
[144] O. Chapelle, P. Haffner, and V. N. Vapnik. Support vector machines forhistogram-based image classification. IEEE Transactions on Neural Networks,1999,10(5):1055-1064.
[145] J.J. Rodriguez， J. Maudes. Boosting Reeombined Weak Classifiers. PatterRecognition Letters,2008,29:1049-1059.
[146] Mattia Marconcini, Gustavo Camps-Valls, and Lorenzo Bruzzone. A CompositeSemisupervised SVM for Classification of Hyperspectral Images. IEEE Trans.Geosci. Remote Sens.2009,6(2):234-238.
[147] Asuncion,A.&Newman,D.J.(2007).UCI Machine Learning Repository
[www.ics.uci.edu/~mlearn/MLRepository.html].Irvine,CA:University ofCalifornia, School of Information and Computer Science.
[148] J.J. Rodriguez， J. Maudes. Boosting Recombined Weak Classifiers. PatterRecognition Letters,2008,29:1049-1059.
[149] Xuchun Li, Lei Wang. AdaBoost with SVM-based component classifiers.Engineering Applications of Artificial Intelligence,2008(21):785–795.
[150] E. Romero, L. Marquez, X. Carreras, Margin maximization with feed-forwardneural networks: a comparative study with SVM and AdaBoost. Neuro-computing,2004,(57):313-344.
[151]何斌，马大予，王运坚，等.visual c++数字图像处理(第二版).北京:人民邮电出版社出版社，2002.