Dim Small Target Detection and Tracking: A Novel Method Based on Temporal Energy Selective Scaling and Trajectory Association

TL;DR

This work addresses the challenge of detecting and tracking dim, small targets in passive optical remote sensing under low SCR by exploiting temporal information across multiple frames. It introduces Temporal Energy Selective Scaling (TESS) to amplify transient disturbances in intensity temporal profiles, followed by a 3D Hough-transform-based trajectory extraction to suppress false alarms, and a trajectory-based tracking scheme using Kalman filtering and cost-based data association. The method demonstrates improved detection performance, reduced false alarms, and real-time potential across diverse sequences, highlighting its practical value for early warning and monitoring in cluttered scenes. The approach offers a principled integration of temporal and spatial trajectory information to enable robust multi-target tracking when traditional single-frame and appearance-based methods falter.

Abstract

The detection and tracking of small targets in passive optical remote sensing (PORS) has broad applications. However, most of the previously proposed methods seldom utilize the abundant temporal features formed by target motion, resulting in poor detection and tracking performance for low signal-to-clutter ratio (SCR) targets. In this article, we analyze the difficulty based on spatial features and the feasibility based on temporal features of realizing effective detection. According to this analysis, we use a multi-frame as a detection unit and propose a detection method based on temporal energy selective scaling (TESS). Specifically, we investigated the composition of intensity temporal profiles (ITPs) formed by pixels on a multi-frame detection unit. For the target-present pixel, the target passing through the pixel will bring a weak transient disturbance on the ITP and introduce a change in the statistical properties of ITP. We use a well-designed function to amplify the transient disturbance, suppress the background and noise components, and output the trajectory of the target on the multi-frame detection unit. Subsequently, to solve the contradiction between the detection rate and the false alarm rate brought by the traditional threshold segmentation, we associate the temporal and spatial features of the output trajectory and propose a trajectory extraction method based on the 3D Hough transform. Finally, we model the trajectory of the target and propose a trajectory-based multi-target tracking method. Compared with the various state-of-the-art detection and tracking methods, experiments in multiple scenarios prove the superiority of our proposed methods.

Figures (16)

• Figure 1: Framework of our proposed detection and tracking method. First, we compute the distance of each ITP in a multi-frame detection unit in a parallel way, obtaining the trajectory position matrix and the occurrence moment matrix. Then, we perform trajectory extraction based on the 3D Hough transform to eliminate false alarm points. After the trajectory extraction, we can directly associate the trajectories to get the final detection results. It is also possible to track the trajectories of the detection unit to get the trajectory tracking result.
• Figure 2: Principles of Optical Imaging.
• Figure 3: SCR reduction after doubling, quadrupling, and octupling the detection distance.
• Figure 4: Probability density functions (PDF) of typical pixel for different scenes. (a) Different scenes. (b) Noise signal of a typical pixel. (c) PDF of noise signals.
• Figure 5: The 2D and 3D simulations of small target. (a) The 2D simulation of a small target (b) The 3D simulation of a small target.