The Research Of Image Processing And Discrimination Of Medicine-farina

With the quick development of computer, the processing technology of digital image has been used widely in industry, it's appear is significative for the improvement of labor productivity. The development of economy in our country is very quick, and the market is flourish in recent years, However, the means of making fake commodity of some illegal people who fabricate and sell fake medicine is increasing master, it is damage the healthy of people all over the country terrifically, so it is very significative for researching an exact and quick method of discrimination for fake medicine.The article find a new method of image processing and automatic discriminating for real and fake medicine-farina in capsule, it adopt Computer Vision and can resolve the difficult question of hard discrimination for fake medicine in market. First,get the picture of the medicine-farina by good camera or microscope, and process them in advance ,then get the medicine-farina from the gray-image and the color-image by methods of gray-segment and color--segment. And analyse their configuration,shape and their color-ingredient. At last give an automatic discriminating method for the pictures which are got by camera, and design the discriminating system. Then calculate some characteristic parameters such as form-diathesis,acreage-volume,girth-volume of every one, Those numbers of characteristic parameters were used as inputs to the neural network, A neural network was set up between the feature parameters and real or false medicine-farina. The results of the experiment show that the average accuracy of the identification of the method is 97.5%.