The Design And Implementation Of Thickness Online Detection Of Chinese Medicine Patch

Traditional production of Chinese medicine patch, low level of automation and industrialization, often uses artificial methods to detect the thickness of the sample patch, labor-intensive, low productivity. The thickness detection based on image sensor has the advantages of non-contact detection, online detection and high accuracy which meet the needs of modern production. The content of this paper is online detected the patch thickness's uniformity and thickness. Among the patch thickness's uniformity, including: As the roller tilt jitter caused thickness of uneven tilt and as the Burkina injustice and the membrane-based inequality caused the thickness of fluctuations uneven.Based on the theory of camera model, designed the thickness detection program, and designed the patch thickness online detection system. Then, analyzed and designed each module of the system. The key to system is the extraction of edge of the patch in the image and the treatment of the feature points.Image processing module implemented the single-pixel accurate continuous positioning patch edge. Image processing has three steps: pretreatment of image, image segmentation and edge extraction. In the pretreatment of image including: read part of the image, gray the image and removal outliers of the image. By comparing the traditional threshold methods: simple threshold method, iterative method, Otsu method, and by testing the classic methods of edge extraction, this paper applies Otsu method for image segmentation and Roberts operator for edge extraction.Access to the edge of the patch in the image, designed the measurement method using the features of the thickness detection program and principles of Hough transform. The intercept of the edge line is the average of Feature points on the edge of the Y-axis direction, and the line slope is obtained with the accuracy of 0.1 degrees by minimizing the scope and then detecting with a smaller resolution. The thickness is calculated by the intercept of the line through the patch edge, the intercept of the reference line at the patch bottom and scale factor between thickness of patch and pixel in the image; thickness of uneven tilt is determined by the line slope of the edge line; the thickness of fluctuations uneven is judged through statistical feature points, and then compared with the experience.Finally, this paper designed the entire process of the system, designed and implemented the software in VC++ environment. The results showed that the thickness detection meet the production demand.