基于ImagePy的水敏纸图像预处理及液滴参数测量

快速获取喷施农药后雾滴在水敏纸表面的沉积分布,对了解农药的田间分布情况有十分重要的意义.针对目前水敏纸图像处理软件缺少对水敏纸图像的旋转等预处理操作,以及将水敏纸图像先转换为灰度图像再进行分割而导致检测精度低的问题,基于ImagePy开发了一种可以快速准确检测雾滴尺寸及分布特性的水敏纸图像处理软件,该软件由一组自定义插件组成,可以对水敏纸图像进行自动检测并旋转,然后对水敏纸图像进行彩色图像分割,进而使用这些插件统计测量雾滴尺寸、分布、总雾滴数和雾滴密度等雾滴参数,并将检测结果保存在Excel表格中以便后续数据处理,通过与DepositScan软件进行比较,验证了本软件的有效性.实验结果表明,基于ImagePy的水敏纸图像预处理及液滴参数测量方法可提取试纸上94.117 6％以上的雾滴,较DepositScan软件准确率提升了6.770 7％.

Image preprocessing and droplet parameter measurement of water- sensitive paper based on ImagePy

It is very important to understand the field distribution of pesticides by quickly obtaining the deposition distribution of droplets on the surface of water-sensitive paper(WSP) after spraying pesticide. At present, the water sensitive paper image processing software lacks the pre-processing operation such as the rotation of the water sensitive paper image, also it has the low detection accuracy problem which is caused by converting the water sensitive paper images to grayscale images before segmentation. To address these shortcomings, a water sensitive paper image processing software based on ImagePy is developed in this paper, by which the droplet size and distribution characteristics of water sensitive paper can be quickly and accurately detected. The software consists of a group of user-defined plug-ins, which can automatically detect and rotate the water sensitive paper images, and then, color image segmentation is done for water sensitive paper images. Further, the droplet parameters such as droplet size, distribution, total droplet number and droplet density can be measured by these plug-ins. Finally, the measure results are saved in Excel for subsequent data processing. The effectiveness of the software is verified by comparison with the DepositScan software.The experimental results show that the method of image preprocessing and droplet parameter measurement based on Imagepy can extract more than 94.1176% of droplets on the test paper, and the accuracy rate is 6.7707% higher than that of Depositscan software.