Rough-Fuzzy Image Feature Discretization for Remote Sensing Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing remote sensing image feature discretization methods based on rough sets struggle with low accuracy in complex data types due to the inability to describe fuzzy components effectively, leading to inefficiencies in edge cloud computing and data inconsistency.
Innovation Solution
A remote sensing image feature discretization method using a rough-fuzzy model that iteratively updates class centers and membership degrees, builds a rough-fuzzy set, and employs a genetic algorithm to select optimal discretization solutions, transforming continuous features into discrete ones.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If rough set based discretization method is used, then preprocessing efficiency is improved, but accuracy in describing fuzzy components deteriorates
Solution Approach 1:
The patent merges rough set theory with fuzzy set theory to create a rough-fuzzy discretization model. This combination allows the system to maintain the efficiency advantages of rough sets while incorporating the ability of fuzzy sets to handle uncertainty and fuzzy components in data, thereby resolving the contradiction between preprocessing efficiency and accuracy in describing fuzzy components
Solution Approach 2:
The patent creates a composite theoretical framework by integrating rough set equivalence relations with fuzzy membership functions. This composite approach combines the computational efficiency of rough sets with the descriptive accuracy of fuzzy sets, enabling effective handling of both crisp and fuzzy data characteristics simultaneously
2Device complexity
If equivalence class of rough set is used, then simplicity of model is maintained, but ability to describe fuzzy components deteriorates
Solution Approach 1:
The patent introduces dynamic membership degrees that can adapt to different data characteristics while maintaining the underlying rough set structure. The fuzzy membership functions allow the model to dynamically adjust its description of data points based on their degree of belonging to different equivalence classes, enhancing adaptability without completely abandoning the simple rough set framework
Solution Approach 2:
The patent uses fuzzy membership functions as an intermediary layer between the rough set equivalence classes and the actual data points. This intermediary allows for gradual transitions and partial memberships, enabling the model to describe fuzzy components effectively while maintaining the structural simplicity of rough set equivalence classes
Data Source
AI summary
Provided is a remote sensing image feature discretization method based on rough-fuzzy model, comprising the following steps: listing the digital number in each band and category of a selected sample in remote sensing images, and building an image information decision table based on the digital number and category; initializing the class center of each category and the membership degree of a sample example relative to the class center; updating the class center of each category and the membership degree of the sample example relative to the class center iteratively, and obtaining the final value of the class center of each category and the final value of the membership degree; building a rough-fuzzy set, computing the mean approximation accuracy of the rough-fuzzy set, discretizing the image information decision table, evaluating the discretization results based on the mean approximation accuracy and a genetic algorithm, and selecting an optimal discretization solution.