Pethybridge Sarah J, Nelson Scot C
School of Integrative Plant Science, Section of Plant Pathology and Plant-Microbe Biology, Cornell University, Geneva, NY 14456.
College of Tropical Agriculture and Human Resources, Department of Plant and Environmental Protection Sciences, University of Hawaii at Manoa, Honolulu, HI 96822.
Plant Dis. 2015 Oct;99(10):1310-1316. doi: 10.1094/PDIS-03-15-0319-RE. Epub 2015 Aug 14.
An interactive, iterative smartphone application was used on color images to distinguish diseased from healthy plant tissues and calculate percentage of disease severity. The user touches the application's display screen to select up to eight different colors that represent healthy tissues. The user then moves a threshold slider until only the symptomatic tissues have been transformed into a blue hue. The pixelated image is then analyzed to calculate the disease percentage. This study reports the accuracy, precision, and robustness of Leaf Doctor using six different diseases with typical lesions of varying severity. Estimates of disease severity from Leaf Doctor were highly accurate (R ≥ 0.79; C ≥ 0.959) compared with estimates obtained from the discipline-standard, Assess. Precision was operationally defined as the ability of a rater to use Leaf Doctor and repeatedly obtain similar percentages of disease severity for the same image. Coefficients of variation were low (0.51 to 14.1%) across all disease datasets but a significant negative relationship was found between the coefficient of variation of estimates and mean disease severity. Other advantages of Leaf Doctor included comparatively less time for image processing, low cost, ease of use, ability to send results by e-mail, and the ability to create realistic standard area diagrams. Leaf Doctor is compatible with iPhone, iPad, and iPod touch and is optimized for iPhone 5. It is available as a free download at the iTunes Store.
一款交互式、迭代式的智能手机应用程序被用于彩色图像，以区分患病植物组织和健康植物组织，并计算病害严重程度的百分比。用户触摸应用程序的显示屏，选择多达八种代表健康组织的不同颜色。然后，用户移动阈值滑块，直到只有有症状的组织被转换为蓝色调。接着对像素化图像进行分析，以计算病害百分比。本研究报告了Leaf Doctor针对六种具有不同严重程度典型病斑的不同病害的准确性、精确性和稳健性。与从学科标准评估工具Assess获得的估计值相比，Leaf Doctor对病害严重程度的估计非常准确（R≥0.79；C≥0.959）。精确性在操作上被定义为评分者使用Leaf Doctor并针对同一图像反复获得相似病害严重程度百分比的能力。所有病害数据集的变异系数都很低（0.51%至14.1%），但在估计值的变异系数与平均病害严重程度之间发现了显著的负相关关系。Leaf Doctor的其他优点包括图像处理时间相对较短、成本低、易于使用、能够通过电子邮件发送结果以及能够创建逼真的标准面积图。Leaf Doctor与iPhone、iPad和iPod touch兼容，并针对iPhone 5进行了优化。它可在iTunes商店免费下载。
