| This research of diseased Pinus trees is of great practical significance for the prevention and control of pine forest diseases.Accurate monitoring of Pinus trees growth,and timely detect and determine the location of diseased pine trees,which can control the spread of the disease as soon as possible.Given the dense forest on the steep slopes of mountainous areas,the traditional artificial recognition method is too laborious and inefficient,and it is not conducive to the dynamic monitoring of Pinus tree diseases and cannot ensure the timeliness of monitoring.Remote sensing technology is widely used for the detection of diseased Pinus trees and the damage of plant canopy can be obtained through remote sensing images.Combined with remote sensing and computer vision technology,for high-resolution near-Earth remote sensing images obtained by UAV remote sensing for ground target recognition.Using machine learning and deep learning algorithms to identify diseased Pinus trees can achieve high-precision recognition results.We firstly introduce the traditional machine learning methods for the detection and identification of diseases in agriculture and forestry,including extracted feature types and recognition accuracy.In comparative experiments,we introduce the deep learning,reinforcement learning,and other methods to detect diseases,and compares the experimental results of this method with traditional machine learning and deep learning.The main contents of this work are as follows:1.Firstly,we introduce the geographical environment of the research area,the disease of Pinus trees,the model and parameters of UAV,and image preprocessing obtained by UAV.We introduce the traditional machine learning methods including support vector machine and back propagation network(BP-net),as well as Alex-net,VGG(Visual Geometry Group)and other deep learning methods to identify diseased Pinus trees.2.A method of diseased Pinus trees identification combined with deep convolutional neural network and AdaBoost algorithm is proposed to solve the problem of low accuracy of traditional machine learning method for identifying diseased Pinus trees.Firstly,the convolutional neural network is used to train the diseased pine model and then use the pre-training model to remove the complex information such as fields,bare soil and shadows,and extract the color and texture features of the diseased pine,healthy pine,shadow areas and other vegetation.The AdaBoost classifier is used to identify the diseased target according to the extracted features in the decision-making layer after the object interference item is removed.The experimental results show that the proposed method has significantly higher recognition accuracy than traditional k-means clustering,support vector machine,AdaBoost algorithm,BP neural network and VGG algorithm.3.This study presents a method for recognizing diseased Pinus trees that combines deep convolutional neural network(DCNN),deep convolutional generative adversarial network(DCGAN),and an AdaBoost classifier.Recognition of diseased Pinus trees in UAV images is beneficial to the dynamic monitoring and control of Pinus tree diseases in large areas.However,the low resolution and complex backgrounds of UAV images limit the accuracy of traditional machine learning methods in recognizing diseased Pinus trees.DCGAN can expand the number of samples of diseased Pinus trees to solve the problem of insufficient training samples.DCNN are used to remove fields,soils,roads,and rocks in images to reduce the impact of complex backgrounds on target recognition.The AdaBoost classifier distinguishes diseased Pinus trees from healthy Pinus trees,identifies shadows and other vegetation in background removal images.Then mathematical morphology is used to remove the small error recognition area and improve the overall recognition accuracy.Experimental results show that the proposed method has better recognition performance than K-means clustering,support vector machine,AdaBoost classifier,backpropagation neural networks,and Inception_v3 networks. |
PDF Full Text Request