研究目的
Investigating the effectiveness of a novel moving target tracking approach based on Convolution Neural Network (CNN) for detecting and tracking moving targets in Video Synthetic Aperture Radar (SAR) systems.
研究成果
The proposed CNN-based moving target tracking approach in video SAR system is effective for detecting and tracking multiple moving targets in heavy clutter environments, providing improved detection and tracking performance.
研究不足
The neural network requires a large number of training samples, which are not readily available from multi-frame SAR image sequences, limiting its application in this field.
1：Experimental Design and Method Selection:
The study employs CNN for shadow detection and tracking of moving targets in SAR imagery. The methodology includes pre-processing of shadow of moving target to form a sequence of sub-aperture SAR images and generate track proposals for CNN inputs.
2：Sample Selection and Data Sources:
The study uses both simulated SAR data and Gotcha data set to validate the effectiveness of the proposed algorithm.
3：List of Experimental Equipment and Materials:
Not explicitly mentioned.
4：Experimental Procedures and Operational Workflow:
The process involves shadow extraction of moving target in the first Video SAR frame, tracking through CNN, and performance analysis through simulation and real data set processing.
5：Data Analysis Methods:
The study uses CNN for feature extraction and tracking, with performance analysis based on simulation results and real data processing.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
