| The injection plays an important role in various medicine products, which is also applied abroad in clinic. However, there are some insoluble impurity particles in injection. If injected into the body following injection, they will be very harmful to human. Recently, people pay more attention to unhealthy reactions caused by injection. At present, pharmaceutical corporations in China use a light inspection method carried out by workers. This method is simple, but the inspection efficiency and repeatability are poor, with omission rates increasing synchronously due to workers’tiredness and workers’judge greatly. In consideration of the situation above, it is necessary to develop an advanced intelligent liquid inspection machine to achieve the online quality inspection of injection products.This paper did some research on the key technologies of intelligent inspection machine for injection impurity inspect. Intelligent inspection machine is a equipment refers to optics, mechanics, electrics, air theory. The development of it involved several subjects and fields, including computer vision, optics, precise mechanical, electrics, image processing, microprocessor control, etc. It is a multi-subject crossing problem. Applied automatic inspection taking over artificial light inspection, the equipment can achieve online automatic detection. This paper studied some key technologies of intelligent inspection machine for injection impurity inspect, the main contents are as follows,1. Injection impurity intelligent inspection machine based on line CCD and algorithm proposed in the paper was developed. And this paper studied this equipment. Special optics system was designed in this paper. This paper chose camera, image acquisition card, lens and some key component according to the demand.2. The mechanical system of injection impurity intelligent inspection machine is designed in this paper. In this paper, mechanical structure has been designed of clamping conveying system of this equipment. And the3-D model has been done of the clamping conveying system, which including clamping platform, delivery system, driving and braking module, and so on. Meanwhile, this paper did static mechanics simulation analysis of the clamping platform in the ANSYS Workbench, which provides references for the following work. Finally, this paper developed an experiment machine used to detect single injection bottle, which verified the feasibility of the design scheme for clamping conveying system.3. After assembling the experiment mechanical system, we debugged the function of this machine, developed the electric system. The information communicate of the experiment machine was achieved by serial communication. Also, we developed VC++program to control the motion of motor. And experiments for injection image acquisition have been taken. The result of experiments showed that the development scheme in the paper can meet the function demand to detect injection impurity.4. Algorithms for injection impurity processing fit this system were proposed in the paper. We used bilateral filtering in image processing, and used threshold segmentation in binary processing. This paper carried on the processing of real injection impurity images obtained from image acquisition experiment of experiment machine, which verified the rationality of the algorithm. The results of experiments showed that the image processing method can detect the impurities in the injection effectively. |