The Studies Of The Rice Quality Inteligent Tester Based On Gb Parameters

In the "high-quality rice standard" of GBT/17891-1999, state bureau of quality and technical supervision increases some important quality indicators of rice shape, such as the head rice rate, the chalkiness degree,the Chalky grain rate, and grain type of rice, etc.However, the above parameters for indicators mainly rely on manual detection method to obtain.Manual detection method has some poor characteristics, such as subjectivity strong, and poor reproducibility, and the determination results of different people are vary widely. In the rice purchase, storage and other processes,the results is difficult to meet the rapid test for quality, objectivity and accuracy requirements.In order to meet the testing of rice quality parameters, this article designs a quality rice intelligent tester based on embedded system.The system is based on ARM920T core S3C2410 platform. We design the rice image acquisition device based on webeye2000. The system designes LCD display circuit and the PS2 keyboard interface circuit.Using embedded linux system, the system completes embedded Linux kernel cutting and transplantation and JPEGLIB transplantation.We have prepared image capture program based on the V4L.In order to obtain the accurately grain image, the system uses six 3W lamps to obtain uniform light source, and the brightness of the light source can be adjusted to make the grain image and background image pixel value with the appropriate gradient.So that the system completes the edge detection of different rice varieties.Using regional labeling technology, the system The system completed a single grain image from the whole image. Based on those, the system completes the grain edge detection.We apply image processing and pattern recognition technology, and detect the GB parameters of rice. To obtain the Chalkiness parameters, the system separate the image twice using OTSU method. The image is divided into foreground and background using the T1.And the background image and chalky image binarization segmentation using the T2, and the system completes the chalkiness parameters calculation. For the grain shape parameters, using the distance formula between two points, the system The system make the coordinate system using the grain image left corner as the coordinate origin. At last,the systen calculates the head rice rate and the shape type.The article lists the corresponding human and system test results,the relevance is 0.99. The systen calculates the head rice rate, the chalkiness degree , the Chalky grain rate, and the shape type, completes a rapid non-destructive testing process.The device provides a more objective detection methods for the detection of the grain GB parameters, and promotes the development of high quality rice Detection.