Explore the vast range of titles available.

What are the main findings? * A velocity-aware multi-feature fusion method was proposed for classifying ground and low-altitude targets using measured X-band pulse-Doppler radar echoes. * By combining echo preprocessing and discriminative feature extraction, the proposed method achieved robust separation of humans, vehicles, and UAVs in cluttered outdoor environments. A velocity-aware multi-feature fusion method was proposed for classifying ground and low-altitude targets using measured X-band pulse-Doppler radar echoes. By combining echo preprocessing and discriminative feature extraction, the proposed method achieved robust separation of humans, vehicles, and UAVs in cluttered outdoor environments. What are the implications of the main findings? * The proposed framework provides a practical approach for radar target classification under complex outdoor clutter conditions. * The study supports the use of measured radar data and multi-feature fusion for ground surveillance and low-altitude target monitoring. The proposed framework provides a practical approach for radar target classification under complex outdoor clutter conditions. The study supports the use of measured radar data and multi-feature fusion for ground surveillance and low-altitude target monitoring. Classification of ground and low-altitude targets in radar remote sensing is challenging because environmental clutter and noise can significantly degrade the discriminability of target echoes, especially under complex outdoor observation conditions. To improve the classification performance for humans, vehicles, and unmanned aerial vehicles (UAVs), this paper proposes a velocity-aware multi-feature fusion method based on measured radar echo data. First, radar echoes are preprocessed using a wavelet-decomposition-based strategy to suppress clutter and noise while preserving useful target information. Then, multiple complementary features, including wavelet packet energy distribution, spectral entropy, spectral standard deviation, temporal standard deviation, amplitude dispersion coefficient, and relative radar cross-section (RCS), are extracted to characterize the target echoes from different perspectives. Considering the influence of target velocity on Doppler distribution and class separability, the measured data are further divided into different velocity intervals for stratified classification. Based on the fused feature vectors, a long short-term memory (LSTM) network is employed to model feature relationships and perform target classification. Experiments conducted on real measured radar echo data demonstrate that the proposed method achieves classification accuracies of 97.82% for UAVs, 96.00% for vehicles, and a mean interval-level accuracy of 96.94%, indicating its effectiveness for ground and low-altitude target classification in cluttered radar remote sensing environments.