Citation: Zhang, X.; Zhang, H.; Chen,
Z.; Li, J. Trunk Borer Identification
Based on Convolutional Neural
Networks. Appl. Sci. 2023, 13, 863.
https://doi.org/10.3390/
app13020863
Academic Editors: Phivos Mylonas,
Katia Lida Kermanidis and
Manolis Maragoudakis
Received: 8 December 2022
Revised: 29 December 2022
Accepted: 4 January 2023
Published: 8 January 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Trunk Borer Identification Based on Convolutional
Neural Networks
Xing Zhang
1,2
, Haiyan Zhang
1,2,
*, Zhibo Chen
1,2
and Juhu Li
1,2
1
School of Information Science and Technology, Beijing Forestry University, Beijing 100083, China
2
Engineering Research Center for Forestry-Oriented Intelligent Information Processing of National Forestry
and Grassland Administration, Beijing 100083, China
* Correspondence: zhyzml@bjfu.edu.cn
Abstract:
The trunk borer is a great danger to forests because of its strong concealment, long lag
and great destructiveness. In order to improve the early monitoring ability of trunk borers, the
representative Agrilus planipennis Fairmaire was selected as the research object. The convolutional
neural network named TrunkNet was designed to identify the activity sounds of Agrilus planipennis
Fairmaire larvae. The activity sounds were recorded as vibration signals in audio form. The detector
was used to collect the activity sounds of Agrilus planipennis Fairmaire larvae in the wood segments
and some typical outdoor noise. The vibration signal pulse duration is short, random and high
energy. TrunkNet was designed to train and identify vibration signals of Agrilus planipennis Fairmaire.
Over the course of the experiment, the test accuracy of TrunkNet was 96.89%, while MobileNet_V2,
ResNet18 and VGGish showed 84.27%, 79.37% and 70.85% accuracy, respectively. TrunkNet based on
the convolutional neural network can provide technical support for the automatic monitoring and
early warning of the stealthy tree trunk borers. The work of this study is limited to a single pest. The
experiment will further focus on the applicability of the network to other pests in the future.
Keywords: convolutional neural networks; trunk borer; vibration signal; voice recognition
1. Introduction
Forestry resources are extremely important for the comprehensive development of
China and the stability of the ecological environment. However, the forest is extremely
vulnerable to the destruction of the trunk borer. Timely detection and early treatment
of pests are undoubtedly the biggest difficulty [
1
], followed by continuous updates and
iterations of monitoring and identification technology for pests, aiming to address this
hidden danger. Over the years, trunk borers have caused serious damage to forests, not
only causing water loss and nutrient loss of trees, but also endangering the growth of the
main trunk of trees [
2
]. Trunk borers generally live in hosts in the early days, nibbling
on branches to obtain nutrients and destroying the tissue structure of trees. Because it
is difficult to find and hard to control, the challenge is posed to the detection of pests.
Traditional pest detection methods make it difficult to find pests which are hidden in the
trunk, and thus, foresters miss the best control period. Therefore, the harm is further
aggravated, resulting in irreparable losses, especially in rare trees. In recent years, with the
application of acoustic technology in pest control, new directions in thinking [
3
,
4
] for the
early identification of pests have been provided.
Sound detection has gradually become a new type of pest detection method [
5
,
6
].
Compared with traditional methods such as spot detection and pheromone trapping, the
method of detecting the activity sound of pests in trees has the advantages of easy conve-
nience, early warning time, high efficiency and low destructiveness [
7
,
8
]. As early as the
1920s, people began to pay attention to the acoustic characteristics of insects, classifying
them as characteristic information of insects and testing them [
9
]. With the development of
technology, the use of electronic devices for food acoustic detection of fruit pests has proved
Appl. Sci. 2023, 13, 863. https://doi.org/10.3390/app13020863 https://www.mdpi.com/journal/applsci
