Research On Visual Analysis Methods Of The Forest Disease And Pest Data

With the continuous development of forestry informationization and intelligent technology,the forestry data acquisition and storage capacity are continuously enhanced.the volume of forestry data will continue to Yrow.and the data formats and types will become more diversified.Based on the visual and visual analysis methods,it can deeply analyze and understand the patterns and laws of forestry data to solve the problems in forestry production and research.and bring,new opportunities and challenges for the development forestry industry.Forests are vulnerable to various natural disasters and human factors.Forest pests and diseases.as the most important natural disasters in forests,pose serious threats to forest resources and cause significant losses to forestry production.There are many difficulties and challenges in the analysis of forest pest and disease data.First.forest pest and disease data is large in volume.complex in structure.multi-level and high-dimensional.and involves space-time attributes.The analysis results of different time and space granularities vary widely.Second,the attributes in the data are not completely isolated,and there are different degrees of association between the attributes.It is difficult to visually present the connections and laws between data by using traditional statistical methods to extract valuable information.Data visualization is an effective way to enhance data cognition using human-perceivable visual symbols,which can help data analysts visually observe and analyze the laws of data.Aiming at the problems in data analysis of forest pests and diseases,this paper takes interactive visual analysis as the core,and carries out analysis and research on forest pest and disease data modeling,visualization and visual analysis scheme design,to provide better management and monitoring for forest pests and diseases research and management personnel,and to provide a more favorable platform for guiding scientific prevention and control of pests and diseases.The main research contents and contributions of this paper are summarized as follows:1.A visual data cleaning method was designed to improve the data quality of forest pest and disease data.To compare the similarity of forest pest and disease text data in the data cleaning process,a text-type data similarity matching algorithm was proposed.Aiming at the characteristics of forest pest and disease data,a visual data cleaning framework was designed to interactively detect,analy ze and clean data to achieve effective control of data quality.2.A clustering data visual analysis method was designed,which can quantitatively assess the similarity of forest pests and diseases in various regions.In the research of visual rendering algorithm.the weighted average ordered treemap layout algorithm was proposed to optimize the treemap to display the ordered hierarchical data in the forest pest and disease data.The clustering edge binding algorithm based on gravitational field was proposed to optimize parallel coordinates to show the distribution characteristics of forest pest and disease cluster data.Based on this,a data clustering visualization scheme for revealing the similarity of forest pests and diseases in various regions was proposed.3.Design of multi-view collaborative visual analysis method based on three models.A multi-view collaborative configurable model was proposed,which can be configured for similar data analysis scenarios.Based on whether the data attributes of different scenarios are consistent,different visual analysis templates are designed to analyze the situation of forest pests and diseases occurrence and control.A hierarchical association interaction model was proposed to guide multiple hierarchical attributes for progressive association interaction analysis.Based on this model,an interactive multi-view collaborative visual analysis method for analyzing the occurrence and development of different pests and diseases in different regions was proposed.A multi-combined multiple linear regression model was proposed to quantitatively describe the linear relationship of multiple combinations between multiple independent variables and single dependent variable.Based on the model and data flow model combined with statistical principles and visualization techniques,the multi-combined multiple linear regression visual analysis method was used to analyze the characteristics of forest pest and disease index and the factors that may cause it to occur.4.A visual analysis prototype system based on forest pest and disease data was designed and implemented.Based on models and methods proposed in this paper,combined with the sequence,geography,disaster level and type of forest pests and diseases,comprehensive consideration of the occurrence and control of different periods and regions,and the influencing factors of forest pests and diseases,to achieve forest pests and diseases multi-angle comprehensive analysis to provide a quick and convenient tool for observing and analyzing forest pests and diseases.5.For the prototype system designed and implemented by this paper,studies were conducted through real forest pest and disease data,and the spatial-temporal and the relationship between multi-dimensional attributes of the data were analyzed to explore the influencing factors of forest pests and diseases,to find out the key impact factor.Relevant user studies and expert assessments were implemented to verify the availability and effectiveness of the proposed models and methods.The research work in this paper combines data mining,visual analysis and mathematical statistics methods to explore new problems and technical means to solve the problems faced by forest pest and disease data analysis and utilization.It assists forest pest and disease research and management personnel to comprehensively grasp the prevention and control of forest pests and diseases,and provides a basis for scientific prevention measures.