Related Experiment Video
Updated: Dec 26, 2025
11:53
Spatial Multiobjective Optimization of Agricultural Conservation Practices using a SWAT Model and an Evolutionary Algorithm
Published on: December 9, 2012
13.3K
Combining Pixel Swapping and Simulated Annealing for Land Cover Mapping
Lijuan Su1, Yue Xu1, Yan Yuan1
1Key Laboratory of Precision Opto-mechatronics Technology Sponsored by Ministry of Education, School of Instrumentation Science and Opto-electronics Engineering, Beihang University, Beijing 100191, China.
Sensors (Basel, Switzerland)
|March 19, 2020
Summary
This study introduces the PSA_MSA algorithm for improved subpixel mapping (SPM) in remote sensing. By combining pixel-swapping with simulated annealing, it overcomes local optima for more accurate land cover classification.
Area of Science:
- Remote Sensing
- Geospatial Analysis
- Image Processing
Background:
- Mixed pixels in low-resolution remote sensing images hinder accurate land cover classification and mapping.
- Subpixel mapping (SPM) techniques, like the pixel-swapping algorithm (PSA), aim to resolve subpixel land cover information but often get stuck in local optima.
Purpose of the Study:
- To develop a novel SPM method that overcomes the local optimum limitation of existing PSA algorithms.
- To enhance the accuracy of subpixel land cover allocation and high-precision mapping.
Main Methods:
- Proposed the PSA_MSA algorithm, integrating the pixel-swapping algorithm (PSA) with a modified simulated annealing (MSA) algorithm.
- The MSA component enables subpixel swapping within a defined range to escape local optima.
- Optimized mixed pixels in a randomized sequence to further boost mapping precision.
Main Results:
- The PSA_MSA algorithm demonstrated superior performance compared to existing PSA-based SPM algorithms.
- The algorithm is particularly effective for remote sensing images exhibiting high spatial autocorrelation.
- Higher proportion errors were found to significantly degrade subpixel mapping accuracy.
Conclusions:
- The PSA_MSA algorithm offers a robust solution for subpixel mapping, achieving global optimum solutions.
- It provides a significant advancement in accurate land cover classification and high-precision mapping from remote sensing data.
- The study highlights the importance of minimizing proportion errors for optimal SPM performance.
