Study On The Foreign Body Detection Of Machine Visionvisual System Based On Image Processing

For medical infusion in the present medical preparations in China have inestimatable importance, is the key to preparation of essential health care and clinical application, application is very wide also very important. Including the medical infusion and small pharmaceutical injection would be more or less influenced by various external factors in the production process and packaging the finished product process, factors including the production environment and human operator, can produce various products unqualified circumstance. For example, the bottle appear insoluble particle; infusion dose bottle is not up to the standard; the bottle body is not perfect wait for a variety of different requirements. Now people are constantly improve the quality of life, in the continuous innovation and development of modern science and technology and industrial technology, consumer quality and quality of transfusion preparation requirements and infusion production quantity demand more and more high. In order to improve the medical infusion preparation product safety, now pharmaceutical production line are essential to the operation of artificial detection filling. But the process efficiency of artificial light inspection machine common too low, also easy to affected workers personal factors, so now the medical infusion manufacturing enterprise in general tend to use automatic only machine vision system to replace the artificial, can guarantee the quality of the products so as to improve the medical security, but also greatly enhance the production efficiency.Key technology of machine vision automatic intelligent lamp checking machine foreign based detection system is the main research of this paper, this paper focuses on the key techniques used in principle and the testing process of full-automatic lamp checking machine foreign body detection system, which uses the machine vision inspection technology to image the medical liquid to be measured the whole series get out, get out of the measured liquid image processing algorithm of special and reasonable, which can identify the insoluble matter and other unqualified phenomenon to be checked in, finally by the machine automatically to judge and remove unqualified liquor. The camera to the test liquid image quality and map the image processing algorithms and techniques like the rationality and validity of the foreign body detection accuracy is the key. Application of automatic light inspection machine can really improve the enterprise liquor detection speed and accuracy, improve the quality and quantity of liquor factory, truly liquid production and testing automation.First of all, this paper introduces the technical status of domestic and foreign pharmaceutical preparation and detection and detection of artificial light inspection; medical supplies liquor detection research background as well as the lamp checking machine liquid detection system and significance of the study at home and abroad about the medical pharmaceutical detection technology research achievements and status. About the principle and content of machine vision system based on image processing technology is proposed, the foreign body for research and overcome difficulties and bottlenecks in the face of different types of.Secondly, designed for automatic inspection machine detection system of different medicine liquid type, its mechanical system on the automatic inspection machine platform; image acquisition system; a detailed analysis of the structure of the computer operating system and industrial control system. Summarizes the key technical links of several systems, including lighting light source; the technology of image acquisition and image recognition technology. In order to meet the inspection machine stability and practicality as well as image acquisition precision, designed specifically for light source lighting lamp checking machine system, selected the suitable for the system of camera, image acquisition card and the right lens, which is able to collect the necessary conditions the measured liquid diagram of high quality, but also for the the back part of image processing and laid an important foundation.Then the image preprocessing algorithm for acquisition of image measurement for liquid. Is the foreign body detection algorithm. Adaptive weighted median filter technology in different degree of the original image to be detected liquid collected to remove noise, using the instantaneous frame difference method and the projection method to eliminate the measured liquid bottle lettering and scratches caused by stationary noise, to enhance the contrast of the image using morphological filtering technology, so as to obtain high quality two the value of the measured liquid image. For smaller and has been moving the foreign body, based on sequence image acquisition to the test liquid on the background suppression methods, utilization and multi frame difference method to image of target with different degree of segmentation.Then, research on the detection algorithm for visible foreign matter measured liquid and the characteristic of the track association of matching. Extraction of target characteristics of medicine image sequences collected, track continuous algorithm uses multi frame difference image sequence of target motion to small foreign body movement to be measured in the presence of liquid, method of target analysis foreign body trajectory linear and the moving small target random noise trajectories measured using interference among them, the measured liquid sequence image collected by the lens processing, to verify this algorithm is effective and reasonable.Finally,combined with the full-automatic lamp checking machine the foreign body detection system, put forward some other image processing algorithms for foreign body detection system is reasonable and effective. The development and application of machine vision foreign current detection technology is presented.