Related Experiment Video
Updated: May 3, 2026
08:27
Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines
Published on: January 5, 2024
1.8K
A method for the evaluation of image quality according to the recognition effectiveness of objects in the optical
Tao Yuan1, Xinqi Zheng1, Xuan Hu1
1School of Land Sciences and Technology, China University of Geosciences, Beijing, China ; Key Laboratory of Land Regulation, Ministry of Land and Resources, Beijing, China.
Plos One
|February 4, 2014
Summary
A new method uses object recognition rates (ORR) to assess optical remote sensing image (ORSI) quality. This approach offers a more application-focused evaluation than traditional metrics.
Area of Science:
- Remote Sensing
- Image Analysis
- Machine Learning
Background:
- Effective image quality assessment (IQA) is crucial for optical remote sensing images (ORSI).
- Traditional IQA methods may not fully capture application-specific quality nuances.
- Developing objective and effective IQA methods is an ongoing challenge.
Purpose of the Study:
- To introduce a novel IQA method for ORSI based on standardized object recognition rate (ORR).
- To model ORSI quality under various imaging conditions.
- To compare the proposed method with conventional IQA indicators.
Main Methods:
- Simulated quality degradation of high-resolution ORSI datasets.
- Application of a machine learning algorithm to determine ORR for a target object.
- Quantitative comparison of ORR-based assessment with existing IQA metrics.
Main Results:
- The original ORSI achieved an ORR of 81.95%.
- Quality-degraded images showed ORR ratios ranging from 64.58% to 73.11% compared to the original.
- The ORR-based method demonstrated superior ability in reflecting image utility for object identification.
Conclusions:
- The ORR-based IQA method provides a more accurate assessment of ORSI quality from an application perspective.
- This approach effectively highlights differences in object identification and information extraction capabilities.
- The study presents a new, objective methodology for ORSI quality assessment using machine learning.
