Detection Method To Quality Of The Seed Metering Device Of The Precise Sowing Based On The Machine Vision

The precise sowing is an important approach which improve the output of the modern agriculture. Detecting the performance of the seed metering device (SMD) play an important role in the precise sowing. This paper mainly researched how to detect the performance of the SMD based on the machine vision. Many aspects included, the equipments of this system, the vision detecting of the SMD, the image processing and so on . The using of the Visual C++ realized the image processing, the achieving of characteristics, the monitoring and controlling of experimental results and other functions. In order to detect the properties of the SMD, the system was established and experimented. As compared to the custom ways, no touch, no hurt and no pollution will be came true in the detecting of the performance of the SMD. Moreover, the system can detect the performance fast and automatic. From the experiment, undoubted, it is predicted that the methods of the vision detecting will be used widely in the futural agriculture engineering. The procedure modules on the prophase image processing and how to achieve the characteristic of the seed were programmed. Through the procedure module of the prophase image processing, the video image was partitioned, sharpened, smoothed and it's background also was removed. And corresponding parameter of the image processing was well chosen, the threshold value of the image was 70.The medium filtering and the method of 8 neighboring regions were adopted during the smoothing and partitioning of the image. After the analysis of the movement of the seed and demarcating of the image, the algorithm of identifying the number of cells of the SMD was deduced. Through the procedure module to achieve the characteristic of the seed, such as the amount, position and the number of cells of the SMD, were all achieved. According to the test method of the precise sowing in the International Standard of the sowing equipment offered the mathematical model about counting the precise sowing properties, including the quality of feed index, the multiples index and the miss index. The rotation speed and the air pressure 's effect on the properties of the SMD also been discussed. The reason of the error was expatiated, and all factors which effected on detecting precision of the system were analyzed.