| Dropping pills are widely used in medicine,tobacco and food industry.In the process of mass production,we need to control the size and the wall thickness of the drop pill.For the measurement of the parameters above,the traditional method uses contact measurement technology to obtain the geometric parameters.The traditional method wastes a lot of time waiting for the drying of dripping pills.In this paper,a geometric parameter on-line detection technique for dropping pills based on machine vision was developed in order to guide the production of dropping pills.In view of the advantages of non-destructive detection,low cost,fast detection speed and high detection accuracy,the detection technology based on machine vision was used.Image acquisition system and image processing unit are main analysis object.The image acquisition system includes the illumination system,the camera and the dropping pill detection platform.According to the requirements of the detection of the geometric parameters of the dropping pills,the appropriate design scheme was selected.In this paper,the factors affecting the accuracy of the dropping pills were analysed,and the feasible scheme was put forward.And it was used to guide the structure design of the dropping pills.And an image acquisition system was set up successfully.The image processing unit applied a variety of image processing techniques,including image grayscale,image filtering,image segmentation,image recognition.According to the actual situation of the images,select the appropriate image processing program,verified by experiments.In this paper,the edge detection of image segmentation was designed in detail.And the adaptability of the traditional Canny operator to the image was improved.A high quality image acquisition system has been set up,and a software algorithm for extracting geometric parameters of drop pills from drop pills has been designed.The experimental results show that the detection system has high detection accuracy. |
PDF Full Text Request