A High-Frame-Rate-Imaging-Based Framework for Moving Point Target Detection in Very Low SNR

Abstract

This article proposes a framework for low-signal-to-noise-ratio (SNR) moving point target detection based on the temporal profile analysis of high-frame-rate image sequences. The main idea is that a weak transient disturbance will appear in the time series of target-present pixels and change the statistical characteristics of the temporal profile in the very short high-frame-rate sampling time. In the framework, the target detection workflow, the temporal mathematical model of the moving point target, the pixel modeling process, and key impact factors of detection ability are presented. First, the temporal mathematical model of pixels affected by a moving point target in high-frame-rate image sequences is presented, which gives guidance for designing target detectors in different situations. Then, a pixel modeling approach is proposed to characterize the temporal behavior of pixels for distinguishing between background and target-present pixels. Meanwhile, the key impact factors, such as the frame rate, target velocity, spatial resolution, SNR, and target size, on the detection ability are studied. Finally, a kernel-method-based target detector is presented to demonstrate the effectiveness of the framework. The experimental results with both simulated and real-world data demonstrate that the approach can efficiently detect extremely dim targets based on high-frame-rate imaging.