A Study for Texture Feature Extraction of High-Resolution Satellite Images Based on a Direction Measure and Gray...

Area of Science:

• Computer Vision
• Image Processing
• Remote Sensing

Background:

• Effective image texture feature extraction is crucial for image analysis.
• Existing methods may not fully capture directional information inherent in textures.

Purpose of the Study:

• To propose a new texture feature extraction method by fusing direction measure and Gray Level Co-occurrence Matrix (GLCM).
• To evaluate the performance of the proposed method in classifying high-resolution remote sensing images.

Main Methods:

• A direction measure statistic was developed based on image texture directionality.
• A fusion algorithm integrated the direction measure with GLCM for texture feature extraction.
• Support Vector Machine (SVM) classifier was used for image classification experiments.

Main Results:

• The fusion algorithm effectively extracts texture features by incorporating directional information.
• Qualitative and quantitative assessments confirmed the superiority of the proposed method.
• Significant improvements in image recognition and classification accuracy were observed.

Conclusions:

• The proposed direction measure and GLCM fusion algorithm offers a robust approach for texture feature extraction.
• This method enhances the classification performance of high-resolution remote sensing images.