Recognition Algorithm And Implementation Of Small Foreign Substance In Infusion Soft Bag

A small amount of insoluble foreign substances will be inevitably mixed with medical medicament in the production process,such as plastic scraps,fibers,hairs and so on.If these foreign particles are input to the body,they will bring about some injures and even some people will lose their precious lives.And nowadays there are three kinds of packaging materials,which are glass bottles,plastic bottles and soft bags.For the foreign substances inspection of the glass bottles and the plastic bottles,it has entered the stage of automation from the artificial detection stage,and there are some mature products both at domestic and abroad.However,for the foreign substances inspection of the infusion soft bags which are used at a wider range and has better performance,there are not yet mature products both at domestic and abroad,so that the detection is still at the artificial stage.But light inspection workers are always fatigue and influenced by objective factors,so that the test results are not stable.Therefore in order to overcome the drawbacks of artificial light inspection and improve the factory automation level and economic benefits of the infusion soft bag in the small object detection,it is of great theoretical significance and application value to study the algorithm of detecting foreign particles in the soft bags.Based on the machine vision,the detection technology of small objects in infusion soft bag is studied in this paper.According to the characteristics of the infusion soft bag and impurities in the bags,a good pretreatment rotating platform and light source scheme and control system are designed,improving the image quality.Finally according to the pictures which are took,the appropriate algorithms are designed and the foreign substances detection of the infusion soft bags are completed.The main content of this paper is:First,in order to lay the theoretical foundation for the following algorithm,the dynamic principle of visible foreign substance detection is introduced,and the equation of motion is established.And then,because the color is an obvious feature of an object,the common color space and color segmentation technology of the image are introduced,and the RGB color space is finally decided to use.Second,the characteristics of the material itself and impurities in the infusion soft bags are analyzed.At the same time,in order to make the foreign particles in motion,several kinds of the rotating platform of experimental pretreatment are designed based on these characteristics,and the final choice is that the stepper motor rotates in a horizontal direction way.Then from two aspects of geometrical and physical optics,LED illumination enhancement scheme is designed,and the final choice is that the infusion soft bag is placed in the inclined position and the LED beads graze into the soft bag from the bottom.And at the same time,the two polarizers are respectively added after the light source and before the camera,improving the image contrast between the foreign particles and the background.Third,according to the actual needs,the Basler's acA4600-10 uc type camera and its supporting C125-1620-5M lens for image acquisition are chose.However,after doing the experiment,tailing phenomenon of the foreign particles are found.Finally choosing the stroboscopic light source and digital controller with a camera together to solve this problem.Fourth,according to the pictures obtained,some suitable algorithms are designed,mainly including three aspects: color,kinematics and morphology,and finally the foreign particles are detected successfully.