Model-Based Identification of Larix sibirica Ledeb. Damage Caused by Erannis jacobsoni Djak. Based on UAV Multispectral Features and Machine Learning

Abstract

While unmanned aerial vehicle (UAV) remote sensing technology has been successfully used in crop vegetation pest monitoring, a new approach to forest pest monitoring that can be replicated still needs to be explored. The aim of this study was to develop a model for identifying the degree of damage to forest trees caused by Erannis jacobsoni Djak. (EJD). By calculating UAV multispectral vegetation indices (VIs) and texture features (TF), the features sensitive to the degree of tree damage were extracted using the successive projections algorithm (SPA) and analysis of variance (ANOVA), and a one-dimensional convolutional neural network (1D-CNN), random forest (RF), and support vector machine (SVM) were used to construct damage degree recognition models. The overall accuracy (OA), Kappa, Macro-Recall (Rmacro), and Macro-F1 score (F1macro) of all models exceeded 0.8, and the best results were obtained for the 1D-CNN based on the vegetation index sensitive feature set (OA: 0.8950, Kappa: 0.8666, Rmacro: 0.8859, F1macro: 0.8839), while the SVM results based on both vegetation indices and texture features exhibited the poorest performance (OA: 0.8450, Kappa: 0.8082, Rmacro: 0.8415, F1macro: 0.8335). The results for the stand damage level identified by the models were generally consistent with the field survey results, but the results of SVMVIs+TF were poor. Overall, the 1D-CNN showed the best recognition performance, followed by the RF and SVM. Therefore, the results of this study can serve as an important and practical reference for the accurate and efficient identification of the damage level of forest trees attacked by EJD and for the scientific management of forest pests.

Potential of Unmanned Aerial Vehicle Red–Green–Blue Images for Detecting Needle Pests: A Case Study with Erannis jacobsoni Djak (Lepidoptera, Geometridae)

“…In this paper, traditional machine learning RF and deep learning CNN were utilised as algorithms for model construction. When using all the samples for modelling, CNN showed better results compared to RF, a finding which is consistent with our previous research and Kumar et al’s findings on using deep learning for diagnosing plant early blight and late blight [ 38 , 65 ]. In addition, we found that the size of the sample had an impact on the model construction process, so we tested its performance, and an optimal model was obtained by changing the sample size.…”

“… where DR denotes the rate of leaf loss and takes values between 0 and 100%, and L h and L d denote the number of healthy and damaged needles, respectively. On this basis, through the experience of visual discrimination in field and the classification criteria in previous studies, the results were classified into pest damage levels based on Table 1 [ 38 ], where the final classification results of 210 larch trees into healthy, mild, moderate, and severe levels are shown.…”

“…The model usually consists of five parts: input, convolution, pooling, dense connection, and output. It can map high-dimensional nonlinear data to a low-dimensional space, realise data dimensionality reduction, effectively reduce the number of parameters in the network, and alleviate the overfitting problem of the model [ 38 ].…”

“…where DR denotes the rate of leaf loss and takes values between 0 and 100%, and Lh and Ld denote the number of healthy and damaged needles, respectively. On this basis, through the experience of visual discrimination in field and the classification criteria in previous studies, the results were classified into pest damage levels based on Table 1 [38], where the final classification results of 210 larch trees into healthy, mild, moderate, and severe levels are shown. A test was conducted with a DJI Phantom 4 multispectral quadcopter drone equipped with an all-in-one imaging system with RGB visible sensors (red, green, and blue channels) and five multispectral sensors (blue, green, red, red-edge, and near-infrared bands).…”

Recognition of Rubber Tree Powdery Mildew Based on UAV Remote Sensing with Different Spatial Resolutions

“…Zhao et al [17] selected 15 potential vegetation indices based on UAV multispectral images and used the minimal redundancy maximal relevance (MRMR) algorithm to select three sensitive spectral features for the construction of an areca yellow leaf disease classification model, ensuring the maximum relevance of the feature while removing the advantage of redundant features. Ma et al [18] used analysis of variance (ANOVA) and successive projection algorithm (SPA) to extract sensitive vegetation index features and texture features of forest trees with different damage levels to improve the generalization ability of the Erannis jacobsoni Djak severity model. In addition, the appropriate choice of data analysis method is very important, as it directly affects the reliability and accuracy of the results.…”