The development of computerized automated diagnostic systems ensures more effective health screening in plants. In this way, the damage caused by diseases can be reduced by early detection. Light reflections from plant leaves are known to carry information about plant health. In the study, healthy and fusarium diseased peppers (capsicum annuum) was detected from the reflections obtained from the pepper leaves with the aid of spectroradiometer. Reflections were taken from four groups of pepper leaves (healthy, fusarium-diseased, mycorrhizal fungus, fusarium-diseased and mycorrhizal fungus) grown in a closed environment at wavelengths between 350 nm and 2500 nm. Pepper disease detection takes place in two stages. In the first step, the feature vector is obtained. In the second step, the feature vectors of the input data are classified. The feature vector consist of the coefficients of wavelet decomposition and the statistical values of these coefficients. Artificial Neural Networks (ANN), Naive Bayes (NB) and K-nearest Neighbor (KNN) were used for classification. In detection the health case of pepper, the average success rates of different classification algorithms for the first two groups (diseased and healthy peppers) were calculated as 100% for KNN, 97.5% for ANN and 90% for NB. Likewise, these rates for the classification of all groups were calculated as 100% for KNN, 88.125% for ANN and 82% for NB. Overall, the results have shown that leaf reflections can be successfully used in disease detection.

计算机化自动诊断系统的开发可确保对植物进行更有效的健康检查。这样，可以通过及早发现减少由疾病引起的损害。已知来自植物叶片的光反射会携带有关植物健康的信息。在这项研究中，借助于分光光度计，从辣椒叶片获得的反射中检测出健康的和镰刀菌病的辣椒（辣椒）。反射是在封闭的环境中以350 nm至2500 nm的波长生长的四组辣椒叶（健康的，枯萎病的，菌根真菌，枯萎病的和菌根的真菌）拍摄的。胡椒病的检测分两个阶段进行。在第一步中，获得特征向量。在第二步中，对输入数据的特征向量进行分类。特征向量由小波分解系数和这些系数的统计值组成。人工神经网络（ANN），朴素贝叶斯（NB）和K近邻（KNN）用于分类。在检测辣椒的健康情况时，针对前两组（病态和健康辣椒），不同分类算法的平均成功率计算为：KNN为100％，ANN为97.5％，NB为90％。同样，将所有组的分类率分别计算为：KNN为100％，ANN为88.125％，NB为82％。总体而言，结果表明叶反射可以成功地用于疾病检测。朴素贝叶斯（NB）和K近邻（KNN）用于分类。在检测辣椒的健康情况时，针对前两组（病态和健康辣椒），不同分类算法的平均成功率计算为：KNN为100％，ANN为97.5％，NB为90％。同样，将所有组的分类率分别计算为：KNN为100％，ANN为88.125％，NB为82％。总体而言，结果表明叶反射可以成功地用于疾病检测。朴素贝叶斯（NB）和K近邻（KNN）用于分类。在检测辣椒的健康情况时，前两组（病态和健康辣椒）的不同分类算法的平均成功率计算为：KNN为100％，ANN为97.5％，NB为90％。同样，将所有组的分类率分别计算为：KNN为100％，ANN为88.125％，NB为82％。总的来说，结果表明叶片反射可以成功地用于疾病检测。计算所有组的分类率：KNN为100％，ANN为88.125％，NB为82％。总体而言，结果表明叶反射可以成功地用于疾病检测。计算所有组的分类率：KNN为100％，ANN为88.125％，NB为82％。总体而言，结果表明叶反射可以成功地用于疾病检测。