Abstract
Statistical analysis of Bactrocera grooming behavior is important for pest control and human health. Based on DeepLabCut, this study proposes a noninvasive and effective method to track the key points of Bactrocera minax and to detect and analyze its grooming behavior. The results are analyzed and calculated automatically by a computer program. Traditional movement tracking methods are invasive; for instance, the use of artificial pheromone may affect the behavior of Bactrocera minax, thus directly affecting the accuracy and reliability of experimental results. Traditional research studies mainly rely on manual work for behavior analysis and statistics. Researchers need to play the video frame by frame and record the time interval of each grooming behavior manually, which is time-consuming, laborious, and inaccurate. So the advantages of automated analysis are obvious. Using the method proposed in this paper, the image data of 94538 frames from 5 adult Bactrocera were analyzed and 14 key points were tracked. The overall tracking accuracy was as high as 96.7%. In the behavior analysis and statistics, the average accuracy rate of the five grooming behavior was all above 96%, and the accuracy rate of the remaining two grooming behavior was over 87%. The experimental results show that the automatic noninvasive method designed in this paper can track many key points of Bactrocera minax with high accuracy and ensure the accuracy of insect behavior recognition and analysis, which greatly reduces the manual observation time and provides a new method for key points tracking and behavior recognition of related insects.
Highlights
Bactrocera minax is one of the most serious pests in citrus [1]
Experimental Results e video of Bactrocera minax collected in the laboratory of oriental Bactrocera minax in Agricultural College of Yangtze University in May 2019 was selected as the evaluation sample. e collected videos were 25 frames, with a resolution of 1920 ∗ 1080. 224 images intercepted from ten videos were marked with “left antennae,” “right antennae,” “left side of the body,” “right side of the body,” “left forefoot,” “right forefoot,” “left midfoot,” “right midfoot,” “left hindfoot,” “right hindfoot,” “left wing,” “right wing,” “head,” and “abdomen” with a total of 3136 data. e training results of ResNet-50 are shown in Figure 10. e validation loss value of the first 100,000 cycle training of the network decreases rapidly, and the loss value reduces to 0.00211
In terms of computer vision, a lot of machine learning image processing models and deep learning models have been developed to accurately classify and identify crop pests [46, 47] and a key points tracking and grooming behavior recognition system of Bactrocera minax has been developed inspired by human behavior estimation [48]. e accuracy of traditional machine vision classification mainly depends on the input features. e extracted features are mainly shape, color, texture, and so on, which are usually made by hand, and the original data are converted into feature vectors [49,50,51]
Summary
Bactrocera minax is one of the most serious pests in citrus [1]. Since the 1960s, citrus diseases and insect pests have been increasing year by year. It has been confirmed that inhibition of grooming behavior increases mortality in insect-pathogen bioassay [3, 4]. Grooming behavior provides effective solutions for Bactrocera minax control. Grooming is a very important activity for healthy survival [5] and is a very common behavior [6]. E grooming behavior of insects is a very important part of their defense mechanism [17]. E role of insect grooming and hygienic activities is gaining recognition in the field of insect pathology [18]. Hygienic behavior has been shown to play a key role in disease prevention in insects [19, 20]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
