GLCM-Assisted Satellite Land Cover Classification for Water Budgeting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for managing water usage in water districts are costly, time-consuming, and prone to errors due to inaccurate pixel classification in satellite imagery, leading to incorrect land cover classification, water budget calculation, and water usage analysis.
Innovation Solution
Employ a system that utilizes a grey level co-occurrence matrix (GLCM) and multiple classification components to enhance pixel classification accuracy, followed by a majority vote mechanism to determine final land cover classification, and incorporates regression techniques for water use forecasting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual surveying is used to determine land cover for water budget calculation, then measurement precision can be maintained, but productivity is severely reduced and loss of time increases
Solution Approach 1:
The patent replaces manual mechanical surveying with automated satellite imagery processing and machine learning-based classification systems. The system uses remote sensing data combined with GLCM texture analysis and multiple classification algorithms to automatically determine land cover types, eliminating the need for field surveys while maintaining or improving accuracy.
Solution Approach 2:
The patent creates detailed digital replicas of land cover conditions through satellite imagery and processing algorithms. By analyzing satellite images to generate land cover classifications, the system produces accurate copies of physical land conditions without requiring physical measurement, thereby improving productivity while maintaining measurement precision.
2Device complexity
If single classification component is used for pixel classification, then device complexity is reduced, but measurement precision of land cover classification deteriorates
Solution Approach 1:
The patent combines multiple classification components (GLCM texture classification, vegetation index classification, and machine learning classification) into an integrated system. Each classification method processes satellite imagery through different analytical approaches, and their results are merged to produce the final land cover classification, improving accuracy while managing complexity through systematic integration.
Solution Approach 2:
The patent divides the classification process into separate independent components: GLCM texture analysis, vegetation index calculation, and machine learning classification. Each component handles a specific aspect of land cover identification, allowing for modular development and maintenance while achieving high overall accuracy through the combination of specialized classifiers.
3Ease of operation
If conventional single-method classification is used, then ease of operation is maintained, but reliability of water budget calculation is reduced
Solution Approach 1:
The system performs self-service by automatically processing satellite imagery through multiple classification algorithms and generating land cover classifications without requiring manual intervention. The automated pipeline handles data acquisition, processing, classification, and validation, maintaining ease of operation while improving reliability through robust multi-method classification and error checking.
Data Source
AI summary
A device includes an image data receiving component, a vegetation index generation component, a GLC matrix component, a plurality of classifying components and a voting component. The image data receiving component receives multiband image data of a geographic region. The vegetation index generation component generates a normalized difference vegetation index based on the received multiband image data. The GLC matrix component generates a grey level co-occurrence matrix image band based on the received multiband image data. The classifying components generate land cover classifications based on the received multiband image data, the normalized difference vegetation index and the grey level co-occurrence matrix image band. The voting component generates a final land cover classification based a majority vote of the land cover classifications.