Health assessment of natural larch forest in arxan guided by forestry remote sensing integrated with canopy feature analysis

Abstract

This work aims to propose a more accurate assessment method for forest health in natural larch pine forests of the Arxan by integrating remote sensing technology with tree crown feature analysis. Currently, forest health assessment of natural Larch pine forests relies mainly on ground surveys, and there is a gap in the application of remote sensing technology in this field. This work introduces deep learning technology and proposes a spectral-Gabor space discrimination and classification model to analyze multi-spectral remote sensing image features. Additionally, quantitative indicators, such as tree crown features, are incorporated into the forest health assessment system. The health status of natural Larch pine forests is evaluated using forest resource survey data. The results show that the health levels of natural Larch pine forests in different areas vary and are closely related to factors such as canopy density, community structure, age group, and slope. Both quantitative and qualitative indicators are used in the analysis. The introduction of this innovative method enhances the accuracy and efficiency of forest health assessment, providing significant support for forest protection and management. In addition, the classification accuracy of the health assessment model suggested that the maximum statistical values of average classification accuracy, average classification effectiveness, overall classification accuracy, and Kappa were 74.19%, 61.91%, 63.18%, and 57.63%, respectively. This demonstrates that the model can accurately identify the health status of natural larch forests. This work can effectively assess the health status of the natural larch forest in the Arxan and provide relevant suggestions based on the assessment results to offer a reference for the sustainable development of the forest system.

Multi-Parameter Health Assessment of Jujube Trees Based on Unmanned Aerial Vehicle Hyperspectral Remote Sensing

“…However, the method proposed in this paper is prone to misclassify healthy and sub-healthy jujube trees when assessing their health, which is likely to be due to two reasons. First, healthy and sub-healthy jujube trees differ less in appearance and spectral characteristics [42,43]. Second, the variability between healthy and sub-healthy jujube tree covariates is not significant; however, jujube tree covariate inversion using remote sensing images is susceptible to the influence of the surrounding environment [44,45], background, and other factors, resulting in errors in crop covariates, which in turn leads to a lack of distinction between the two types of jujube trees.…”

Susceptibility Mapping of Unhealthy Trees in Jiuzhaigou Valley Biosphere Reserve

“…Although machine learning algorithms can be effective for monitoring forest species [27][28][29], there may be some limitations in terms of vegetation health. Vegetation indices for forest health mapping actually correspond to the proportion of unhealthy trees within sample plots using medium-resolution (2 m-30 m) remote sensing images [17,[30][31][32][33][34]. In areas where the probability of unhealthy tree occurrence is low, such sample data is difficult to obtain, which is because a single pixel typically corresponds to approximately 100 square meters in medium-resolution image (i.e., Sentinel-2 image).…”

An efficient healthcare system by cloud computing and clustering-based hybrid machine learning algorithm

“…As it is, academics and data scientists face a problem when attempting to make sense of this massive amount of temporal data. The machine learning method has a simple implementation [24]. The data collected may subsequently be utilized to develop algorithms or classification models for stress vulnerability prediction [25].…”