TY - JOUR A2 - Decherchi, Sergio AU - Xu，哲AU - Guo, Xi AU - Zhu, Anfan AU - He, Xiaolin AU - Zhao, Xiaomin AU - Han, Yi AU - Subedi，Roshan PY - 2020 DA - 2020/08/28 TI - Using Deep Convolutional Neural Networks for Image-Based Diagnosis of nutrition deficiency in Rice SP - 7307252 VL - 2020 AB -水稻植株营养缺乏的症状经常出现在叶片上。因此，叶片颜色和形状可用于诊断水稻营养缺乏。图像分类是一种高效、快速的诊断方法。深度卷积神经网络(DCNNs)已被证明在图像分类中是有效的，但其用于识别水稻营养缺陷的研究却很少受到关注。在本研究中，我们探索了不同dcnn诊断水稻营养缺乏的准确性。通过水培试验共获得植物叶片照片1818张，覆盖全营养和10个营养缺乏等级。这些照片按照3:1:1的比例分为训练集、验证集和测试集。对四种最先进的dcnn进行了微调评估:inction -v3、50层ResNet、NasNet-Large和DenseNet(121层)。所有dcnn的验证和测试精度均在90%以上，其中DenseNet121表现最好(验证精度= 98.62±0.57%; test accuracy = 97.44 ± 0.57%). The performance of the DCNNs was validated by comparison to color feature with support vector machine and histogram of oriented gradient with support vector machine. This study demonstrates that DCNNs provide an effective approach to diagnose nutrient deficiencies in rice. SN - 1687-5265 UR - https://doi.org/10.1155/2020/7307252 DO - 10.1155/2020/7307252 JF - Computational Intelligence and Neuroscience PB - Hindawi KW - ER -