Multifractal and its Correlation Features Based Target Classification

This study combines multifractal and correlation analyses to classify aircraft via low-resolution radar echoes. Multifractal correlation quantifies spatial dependencies between singularity indices, while defined parameters capture multiscale distributions and spatial interactions. An SVM trained on combined features achieves 99% accuracy, outperforming standalone methods. Real-data tests confirm robustness to parameter variations and operational constraints. The method enhances classification by synergizing macro-statistical and spatial-correlation features, offering an efficient solution for resource-limited radar systems.