China was one of the earliest countries to set up a system to continuously inventory natural forest resources. From the beginning of the 1970s until today, seven forest resource inventories have been carried out. This research summarizes the progress of forest continuous inventories and analyzes the existing deficiencies of China's forest continuous inventory system and forest management plan inventory. As stated above, this research offers corresponding countermeasures and suggestions: establishing a sample plot system for comprehensive national forest inventory and monitoring with each province's continuous forest inventory based on the foundation of the national sample plot system, able to develop the province as a subset of the overall province-level forest resource inventory according to the actual conditions in each province. Through annual multi-resource/multi-benefit surveying of the forests, the monitoring of forest amounts, quality, functions and benefits will be assisted in its entirety. The further integration of the forest continuous inventory and the forest management plan inventory is also discussed. This research also proposes the varied probability sampling method with sub-compartments as the basic sampling unit (or combinations of sub-compartments). This will also satisfy the requirements of ecological inventory by region.
We studied the use of self-attention mechanism networks (SAN) and convolutional neural networks (CNNs) for forest tree species classification using unmanned aerial vehicle (UAV) remote sensing imagery in Dongtai Forest Farm, Jiangsu Province, China. We trained and validated representative CNN models, such as ResNet and ConvNeXt, as well as the SAN model, which incorporates Transformer models such as Swin Transformer and Vision Transformer (ViT). Our goal was to compare and evaluate the performance and accuracy of these networks when used in parallel. Due to various factors, such as noise, motion blur, and atmospheric scattering, the quality of low-altitude aerial images may be compromised, resulting in indistinct tree crown edges and deficient texture. To address these issues, we adopted Real-ESRGAN technology for image super-resolution reconstruction. Our results showed that the image dataset after reconstruction improved classification accuracy for both the CNN and Transformer models. The final classification accuracies, validated by ResNet, ConvNeXt, ViT, and Swin Transformer, were 96.71%, 98.70%, 97.88%, and 98.59%, respectively, with corresponding improvements of 1.39%, 1.53%, 0.47%, and 1.18%. Our study highlights the potential benefits of Transformer and CNN for forest tree species classification and the importance of addressing the image quality degradation issues in low-altitude aerial images.
This paper focuses on the estimation of green land variation with similarity theory by using two temporal spatial data in Shenzhen. The location, shape and areas of green land units have been used as the similarity elements. Thus the similarity coefficients can be defined. The ratio of overlapping number of green patches to the intersecting number of green patches indicates the location variation of green land. The ratio of minimum to maximum shape index of green land indicates the shape variation of green land. With the same method, the areas variation coefficient has also been obtained. Combined with the analytic hierarchy process, the weights of three similarity elements coefficients can be decided. By comparison with the distinguished threshold the estimation of different green land variation has been made. The result shows that the estimation of green land variation based on similarity theory is feasible. The research of this paper has also provided effective method for the further assessment of green land development in Shenzhen special economic zone.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.