Real-Time Detection of Moving Target Using Through-Wall Radar Based on GPU

Through-Wall radar plays an important role in detecting moving targets behind walls in counter-terrorism and urban-sensing. While there has been a lot of research on moving target detection algorithm design, most of them suffer from high complexity in implementation for real-time detection, and thus limiting the practical applications. This paper proposes a real-time detection technology of moving target by through-wall radar based on GPU. This method includes multi-stream asynchronous parallel technology of the Displaced Phase Center Antenna based on GPU for clutter suppression module, block-thread parallel technology of the Back-Projection based on GPU for imaging module, and inter-block parallel and intra-block reduction technology of 2D Constant False Alarm Rate based on GPU for detection module. Based on CPU-GPU dual devices in through-wall radar system, this method can output the detection result per second, which meets the needs of real-time detection of moving targets in practical applications.