基于图像处理的农药在线混合均匀性分析

Analysis of Pesticide In-line-mixing Uniformity Based on Image Processing Technology

农药喷洒是进行病虫防治,提高农业产量的有效途径,但由于施药技术落后,农药浪费以及环境污染问题日趋严重,为了提高农药使用效率,农药在线混合均匀性值得研究。由于水和含有荧光物质的农药亮度差异比较明显,利用图像处理方法研究农药的在线混合均匀性。搭建农药在线混合图像采集系统,通过差值预处理并选择出面积百分比最大的图像,使用了轴向和径向像素分析方法和均方根误差分析方法对混合均匀性进行定量分析。在此基础上基于Matlab编写了数据分析软件,并对三种不同性质农药的混合均匀性进行分析,其中水溶性农药均方根误差百分比(β)为13. 389%,高于乳化性农药(9. 280%)和脂溶性农药(9.996%),说明相对于乳化性和脂溶性农药,水溶性农药在线混合均匀性较好。且通过对浓度一定但流量不同的7组试验利用该方法对脂溶性农药流体图像进行了数据处理,发现当流量从1 390 m L/min减小到880m L/min时,均方根误差百分比(β)从11.029%减小到8.737%,均匀性逐渐变差。结果表明,利用图像处理方法研究农药在线混合均匀性是可行的,可以方便、直观地分析混药器结构及性能参数对混合效果的影响,为农药在线混合的图像处理方法及混合效果定量分析奠定基础。

Pesticide spraying is an effective way to control pest damage and increase crop yields. Because of the behindhand pesticide spraying technology,the problems of pesticides wast and environmental pollution are getting worse and worse. In order to improve the use efficiency of pesticide,it is worthwhile to study the mixing uniformity of pesticide online. Because there were remarkable differences between water and pesticides containing fluorescent substance,image processing technology was adopted to analyze the uniformity of pesticide on-line-mixing. In this study,an online mixed image collection system for pesticides was set up. The image with maximum percentage of pesticides area in the ROI( region of interest) was chosen after subtraction preprocessing of all acquired images,then the axial and radial pixel analysis method and RMSE statistic were used to quantify the uniformity index. Additionally,the data analysis software was written by Matlab,and the mixing uniformity of 3 different kinds of pesticide were analyzed. The RMSE percentage( β) of water-soluble pesticides was 13.389%,which was higher than 9.280% of emulsified pesticides and 9. 996% of liposoluble pesticides,indicating that the mixing uniformity of water-soluble pesticides was better than that of emulsified and liposoluble pesticides. Experiments on 7 groups with finit concentration but different quantity of flow were carried out data processing on liposoluble pesticide images. When the flow rate decreased from 1 390 m L/min to 880 m L/min,the RMSE percentage dropped from 11.029% to 8.737%,and uniformity became worse gradually. The results revealed that image processing technology was feasible for analyzing uniformity of pesticide on-line-mixing. It could be easily and intuitively adopted to analyze the influence of chemical mixer structure and performance parameters on mixing effect,thus laying a foundation for pesticide on-linemixing image processing,and quantitative analysis of pesticide on-line-mixing effect.