摘要
为了提高喷雾装置的自动对靶精度,实现流量控制的实时自动化调节过程,提出了一种基于二值化图像分割和微波测距的自动对靶喷雾控制系统。该系统利用植物叶间灰度的动态阈值来调控电磁阀,实现了实时变流量喷雾过程。为了提高微波测距的精度和图像二值化过程的效率及鲁棒性,使用了神经网络算法对测距误差进行修正,采用了动态阈值方法对图像的二值化进行迭代计算。最后,对实时自动对靶喷雾控制系统进行了测试,由测试结果可以看出:采用神经网络和动态阈值二值化方法可以大大提升控制系统的精度,提高了喷雾的效率和准确性,为现代化农药喷雾技术的研究提供了理论参考。
In order to improve the accuracy of automatic spraying device for target,realizing the real- time automatic flow control regulation process,proposed a two value automatic target sprayer control system of image segmentation and microwave based on distance measurement,the system uses dynamic threshold of plant leaf between gray to control electromagnetic valve,realizes the real- time variable flow spray process. In order to improve the microwave ranging precision and the two value image efficiency and robustness of the process,the use of neural network algorithm for correction of range error,using the image of two value iteration dynamic threshold method. Finally on the real- time automatic target sprayer control system was tested,by the test results can be seen,using neural network and dynamic threshold value of the two methods can greatly enhance the precision of the control system,improve the efficiency and accuracy of spray,and provides a theoretical reference for the study of modern pesticide spraying technology.
出处
《农机化研究》
北大核心
2016年第2期67-71,共5页
Journal of Agricultural Mechanization Research
基金
国家自然科学基金青年基金项目(51305152)
关键词
自动对靶喷雾
微波测距
神经网络
动态阈值
二值化
automatic target spray
microwave measurement
neural network
dynamic threshold
binaryzation
