Related Experiment Video
Updated: Oct 17, 2025
08:27
Author Spotlight: Efficient Image Recognition Using Directional Gradient Histogram Technique and Support Vector Machines
Published on: January 5, 2024
1.3K
Region Matching of SAR Images Using Blocks for Target Recognition
Chao Shan1, Minggao Li1, Zihao Chen1
1Department of Special Operations Medicine, The Sixth Medical Center of Chinese PLA General Hospital, Beijing 100048, China.
Computational Intelligence and Neuroscience
|October 11, 2021
Summary
This study introduces a new synthetic aperture radar (SAR) target recognition method. By analyzing image blocks with monogenic signals and sparse representation-based classification (SRC), it improves target identification accuracy.
Area of Science:
- Electrical Engineering
- Computer Science
- Remote Sensing
Background:
- Synthetic Aperture Radar (SAR) systems generate high-resolution imagery crucial for target recognition.
- Accurate target recognition in SAR images is challenging due to noise, varying conditions, and complex target signatures.
Purpose of the Study:
- To propose a novel SAR target recognition method utilizing image blocking and feature extraction.
- To enhance the accuracy and robustness of SAR target identification.
Main Methods:
- The proposed method divides SAR images into four blocks for independent analysis.
- Monogenic signal analysis is used to extract feature vectors describing time-frequency distribution and local details.
- Sparse Representation-based Classification (SRC) is applied for classifying feature vectors and generating reconstruction errors.
- A random weight matrix is employed for linear fusion of feature vectors, followed by statistical analysis for final decision-making.
Main Results:
- The method was evaluated on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset.
- Experimental results demonstrate the effectiveness and validity of the proposed SAR target recognition approach.
Conclusions:
- The image blocking and monogenic signal-based feature extraction combined with SRC offers a promising approach for SAR target recognition.
- The statistical fusion strategy contributes to improved classification performance.
