Abstract
Automatic tree identification and position using high-resolution remote sensing images are critical for ecological garden planning, management, and large-scale environmental quality detection. However, existing single-tree detection methods have a high rate of misdetection in forests not only due to the similarity of background and crown colors but also because light and shadow caused abnormal crown shapes, resulting in a high rate of misdetections and missed detection. This article uses urban plantations as the primary research sample. In conjunction with the most recent deep learning method for object detection, a single-tree detection method based on the lite fourth edition of you only look once (YOLOv4-Lite) was proposed. YOLOv4’s object detection framework has been simplified, and the MobileNetv3 convolutional neural network is used as the primary feature extractor to reduce the number of parameters. Data enhancement is performed for categories with fewer single-tree samples, and the loss function is optimized using focal loss. The YOLOv4-Lite method is used to detect single trees on campus, in an orchard, and an economic plantation. Not only is the YOLOv4-Lite method compared to traditional methods such as the local maximum value method and the watershed method, where it outperforms them by nearly 46.1%, but also to novel methods such as the Chan-Vese model and the template matching method, where it outperforms them by nearly 26.4%. The experimental results for single-tree detection demonstrate that the YOLOv4-Lite method improves accuracy and robustness by nearly 36.2%. Our work establishes a reference for the application of YOLOv4-Lite in additional agricultural and plantation products.
Highlights
The smallest tree entity that makes up a terrestrial ecosystem is a single tree
Single-tree extraction from remote sensing images is a critical technology for efficiently constructing a single-tree database built on the foundation of single-tree detection from remote sensing images
If a tree has a large number of branches, multiple maximums will occur, and the local maximum method will result in a high rate of missed and error detection
Summary
The smallest tree entity that makes up a terrestrial ecosystem is a single tree. Single-tree detection and positioning are critical components of precision forestry (Dimitrios and Azadeh, 2021; Dimitrios et al, 2021). Some individual forestry fields in China have implemented intensive management at the single-tree level, such as managing ancient and famous trees, managing female willow trees precisely in Beijing, and managing fruit trees (Xiao et al, 2021) These are primarily accomplished through traditional ground surveys, which are inefficient and require a significant amount of time. 2) Regression-based deep learning object detection algorithms are represented as you only look once (YOLO) (Chaitanya et al, 2020) The former predicts the speeds of between 7 and 18 frames per second, which is too time-consuming. This study performs the urban single-tree detection and positioning using the deep learning network model YOLOv4 (Richey and Shirvaikar, 2021). This model has a high detection speed and is capable of multi-object detection. A YOLOv4-Lite single-tree detection method is proposed to further integrate the YOLOv4 network model (Meneghetti et al, 2021), simplify the entire feature extraction network, optimize loss, and enhance sample data
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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 call
