Related Experiment Video
Updated: May 2, 2026
06:17
Analysis of Multidimensional Microscopy Data Using Cell-ACDC
Published on: November 7, 2025
776
Developing image processing meta-algorithms with data mining of multiple metrics
Kelvin Leung1, Alexandre Cunha2, A W Toga3
1Intel Corporation, 3600 Julliette Ln., Mail Stop SC12-301, Santa Clara, CA 95054, USA ; UCLA Computer Science Department, Los Angeles, CA 90095-1596, USA.
Computational and Mathematical Methods in Medicine
|March 22, 2014
Summary
This study introduces automated mining of multiple image metrics to enhance image processing. This novel approach improves robustness and validation by selecting optimal results from various image measures.
Area of Science:
- Computer Vision
- Image Processing
- Data Mining
Background:
- Image processing commonly employs single metrics for evaluation.
- A need exists for more comprehensive and automated methods to assess image processing outcomes.
Purpose of the Study:
- To propose and demonstrate a novel approach for automated mining of multiple image metric values.
- To extend image processing methods by integrating meta-algorithms for result evaluation.
Main Methods:
- Utilizing a battery of image metrics, including intensity-based and statistical measures.
- Developing meta-algorithms for automated evaluation and selection of image processing results.
- Focusing on image registration as a key application problem.
Main Results:
- Demonstrated the feasibility of automated mining of multiple image metric values.
- Showcased the ability of meta-algorithms to select the best image processing results.
- Identified improved robustness and validation as key benefits.
Conclusions:
- Automated mining of multiple image metrics offers significant advantages for image processing.
- This approach enhances the reliability and accuracy of image processing tasks, particularly in registration.
