Infrared small target detection is a critical task in remote sensing, where targets typically span only a few pixels and are easily submerged by complex background clutter. Traditional convolutional networks predominantly rely on data-driven feature learning or computationally expensive global attention mechanisms. In this paper, we propose the Anomaly-Driven Gated Fusion Network (ADGFNet), a lightweight end-to-end framework that explicitly embeds the statistical characteristics into the feature extraction and fusion pipeline. The core of ADGFNet comprises two synergistic modules: Local Anomaly Block (LAB), which decomposes anomaly perception into a Local Contrast Branch capturing multi-scale intensity saliency and a Gradient Prior Branch extracting boundary discontinuities via Sobel operators; and Anomaly-Guided Feature Pyramid Neck (AG-FPN), which reuses the spatial anomaly maps generated by the LAB as guidance gates to perform spatially selective cross-scale feature fusion without requiring additional learnable spatial-attention modules. Comprehensive evaluations on the NUDT-SIRST and IRSTD-1K benchmarks demonstrate that ADGFNet achieves a superior accuracy–efficiency trade-off, attaining the best nIoU and the lowest false alarm rates with only 0.58 M parameters. With TensorRT FP16 acceleration, ADGFNet and its ultra-lightweight variant, ADGFNet-Lite (0.12 M parameters), enable real-time inference without appreciable loss of detection accuracy on an NVIDIA Jetson AGX Xavier.

As a widely used display device, the effective display area of a digital micromirror device (DMD) is limited by its micromirror count and pitch, which cannot meet large-format target-plane display requirements. This paper proposes a large-format scene simulation system based on multi-DMD optical stitching, and uses an optical relay to overcome the inability to directly tile DMDs because of their package frames. By constructing a DMD display architecture comprising an illumination module, a relay module, a stitching module, and a projection module, the system quadruples the effective display area relative to a single DMD without sacrificing frame rate. Design results show that the system achieves an MTF greater than 0.5 at 50 lp/mm, with near-diffraction-limited performance; the RMS spot radius is less than 10 $\mathsf{\mu }$m. All key indicators meet application requirements.

This study proposed a large-target-surface and high-frame-rate display method using multiple Digital Micromirror Devices (DMDs) for high-resolution, high-frame-rate aerospace applications. DMDs offer high frame rates and contrast ratios, but their surface size is constrained. By employing Pulse Width Modulation (PWM) with synchronization signals for grayscale modulation and synchronizing multiple DMDs, this method achieved a target surface four times larger than a single DMD at 400 Hz frame rate, with synchronization errors below 10 ns. This enhances simulation efficiency and provides an effective infrared scene simulation solution.