Related Experiment Video
Updated: Apr 18, 2026
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
43.9K
Medical image retrieval and analysis by Markov random fields and multi-scale fractal dimension
André Ricardo Backes1, Leandro Cavaleri Gerhardinger, João do Espírito Santo Batista Neto
1Computer Science Department, Federal University of Uberlândia, Av. João Naves de Ávila, 2121-Santa Mônica, Uberlândia MG, 38408-100, Brazil.
Physics in Medicine and Biology
|January 15, 2015
Summary
This study introduces a novel image retrieval signature combining texture and complexity. This new method accurately identifies images, outperforming traditional texture descriptors in pattern recognition and retrieval tasks.
Area of Science:
- Computer Vision
- Image Analysis
- Pattern Recognition
Background:
- Content-based Image Retrieval (CBIR) systems commonly use color, shape, and texture for image analysis.
- Texture is a highly relevant attribute for identifying diverse image types.
- Existing methods often rely solely on texture, limiting retrieval capabilities.
Purpose of the Study:
- To introduce a novel signature for image analysis and retrieval.
- To combine texture segmentation with object complexity estimation.
- To evaluate the proposed method's performance in pattern recognition and image retrieval.
Main Methods:
- Texture segmentation modeled as a Markov Random Field process.
- Estimation of region complexity using Multi-scale Fractal Dimension.
- Experimental validation on an MRI database.
Main Results:
- The proposed method demonstrates high accuracy in image retrieval and pattern recognition.
- The novel signature effectively combines texture and complexity for enhanced image analysis.
- Performance is shown to be sensitive to the level of complexity alteration.
Conclusions:
- The novel texture-complexity signature offers improved performance over traditional texture descriptors.
- The method is effective for both pattern recognition and content-based image retrieval.
- The integration of fractal dimension for complexity estimation enhances image analysis capabilities.
