Related Experiment Video
Updated: Jul 7, 2026
08:27
Image Recognition and Parameter Analysis of Concrete Vibration State Based on Support Vector Machine
Published on: January 5, 2024
Rotation and gray-scale transform-invariant texture classification using spiral resampling, subband decomposition,
Summary
This study introduces a novel texture classification algorithm. The algorithm achieves a 95.14% correct classification rate for 16 textures, demonstrating invariance to rotation and grayscale changes.
Area of Science:
- Computer Vision
- Image Processing
- Machine Learning
Background:
- Texture classification is crucial for image analysis.
- Existing methods often struggle with variations in rotation and grayscale.
- Developing robust algorithms is essential for practical applications.
Purpose of the Study:
- To propose a new texture classification algorithm.
- To achieve invariance to rotation and grayscale transformations.
- To improve the accuracy and reliability of texture recognition.
Main Methods:
- Converting 2D texture images to 1D signals via spiral resampling.
- Decomposing signals into subbands using a quadrature mirror filter (QMF) bank.
- Utilizing high-order autocorrelation functions as features and modeling them with a hidden Markov model (HMM).
Main Results:
- The proposed algorithm demonstrated high performance in simulations.
- Achieved a correct classification rate of 95.14% for 16 different texture types.
- Successfully showed invariance to rotation and grayscale variations.
Conclusions:
- The developed algorithm offers a robust approach to texture classification.
- The combination of spiral resampling, QMF, and HMM is effective.
- The method shows significant potential for various image analysis tasks.
