低空目标探测数据集

数据主编：田彪(中山大学) LSS-HSR-L：全息凝视雷达低空目标探测识别数据集，本数据集是由L波段全息凝视雷达采集的低空目标数据，包含多普勒瀑布图和航迹数据，专门针对低空目标探测与识别任务设计。数据集采集地点涵盖深圳、长沙、重庆等城市、机场、郊区等多种场景。数据集采集到的目标涵盖多种类型，共9类，按大类别可分为四类：第一类是旋翼无人机，包括大疆Air3、大疆Mini3 Pro、大疆Mavic 3e和大疆Phantom 4 RTK；第二类是生物目标，包括雀类小型鸟、鸟群和大型迁徙鸟；第三类是地面固定目标，包括地面固定旋转目标；第四类是移动车辆目标，即汽车。 详细使用说明请参考：数据集及使用说明.zip 本数据集引用格式: 田彪, 陈俊彦, 万延煜, 等. 全息凝视雷达低空目标探测数据集及多特征识别方法[J]. 雷达学报（中英文）, 待出版. doi: 10.12000/JR25212. TIAN Biao, CHEN Junyan, WAN Yanyu, et al. Low-altitude target dataset and multi-feature recognition method for holographic staring radar[J]. Journal of Radars, in press. doi: 10.12000/JR25212. 发布时间：2026年5月15日

Data Editor: Chen Xiaolong (Naval Aviation University)
Dataset Introduction: The Digital Array Comprehensive Detection Radar Low-Slow-Small Target Detection Dataset (LSS-DAUR-1.0) contains 154 collected items of Doppler complex data (TD) and point track data (TR) covering 6 types of targets (passenger ship, speedboat, helicopter, rotary-wing UAV, bird, fixed-wing UAV), which can support the research on maritime typical target detection, classification and recognition of digital array radars.

1. Data Acquisition Process
The data acquisition process mainly includes: set radar parameters → detect targets → collect echo signal data → record target information → determine the range cell where the target is located → extract target Doppler data → extract target track data.

2. Target Categories
The collected typical sea and air targets include 6 categories: passenger ship, speedboat, helicopter, rotary-wing UAV, bird, and fixed-wing UAV.

3. Target Doppler Complex Data (TD Data)
Echo data of the range cell where the target is located is intercepted based on the target range. Using the collected measured data, the low-slow-small TD dataset for digital array comprehensive detection radar is constructed. Specifically, there are 10 sets of passenger ship data, 11 sets of speedboat data, 10 sets of helicopter data, 18 sets of rotary-wing UAV data, 17 sets of bird data, and 11 sets of fixed-wing UAV data, totaling 77 sets. The file structure of the TD dataset is shown in Figure 3.
The naming convention for target TD data is: start_acquisition_time_DAUR_TD_target_type_serial_number_target_batch_number.mat. For example, the file name "20231207093748_DAUR_TD_Passenger Ship_01_2619.mat", where "20231207" represents the data collection date, "093748" represents the start acquisition time at 09:37:48, "DAUR" stands for digital array comprehensive detection radar, "TD" represents target Doppler spectrum complex data, "Passenger Ship_01" indicates the target type is passenger ship with serial number 01, and "2619" represents the target track batch number.

4. Track Data (TR Data)
Track data within the echo data time period is extracted to construct the low-slow-small TR dataset for digital array comprehensive detection radar. There are 10 sets of passenger ship data, 11 sets of speedboat data, 10 sets of helicopter data, 18 sets of rotary-wing UAV data, 17 sets of bird data, and 11 sets of fixed-wing UAV data, totaling 77 sets, forming the TR dataset. The file structure of the TR dataset is shown in Figure 4.
The TR data and TD data share the same time stamp and batch number, which are different dimension data of the same target in the same time period. The naming convention for target TR data is: start_acquisition_time_DAUR_TR_target_type_serial_number_target_batch_number.mat. For example, the file name "20231207093748_DAUR_TR_Passenger Ship_01_2619.mat", where "20231207" represents the data collection date, "093748" represents the start acquisition time at 09:37:48, "DAUR" stands for digital array comprehensive detection radar, "TR" represents range-Doppler spectrum complex data, "Passenger Ship_01" indicates the target type is passenger ship with serial number 01, and "2619" represents the target track batch number.

For detailed usage instructions, please refer to: LSS-DAUR-1.0 数字阵泛探雷达低慢小探测数据集使用说明书.pdf

Relevant papers on "Low-Slow-Small Target Detection Dataset"
[1] Chen X L, Chen W S, Rao Y H, et al. Progress and Prospect of Radar Detection and Recognition Technology for Bird and UAV Targets. Journal of Radars, 2020, 9(5): 803-827.
[2] Chen X L, Yuan W, Du X L, et al. Multi-Band FMCW Radar LSS Detection Dataset (LSS-FMCWR-1.0) and High-Resolution Micro-Motion Feature Extraction Method. Journal of Radars, 2024, 13(3): 539–553.
[3] Chen X L, Rao G L, Guan J, et al. Passive Radar LSS Detection Dataset (LSS-PR-1.0) and Multi-Domain Feature Extraction and Analysis Method. Journal of Radars, 2025, 14(2): 249–268.
[4] Chen X L, Yuan W, Du X L, et al. Multi-Band Multi-Angle FMCW Radar LSS Detection Dataset (LSS-FMCWR-2.0) and Feature Fusion Classification Method. Journal of Radars (Chinese & English), 2025, 14(5): 1276–1293.

Release Date: November 5, 2025