Infrared small target detection can be classified into data-driven methods and model-driven methods. The data-driven methods rely heavily on computational resources and labeled training samples. In contrast, the model-driven small target detection algorithms typically require fewer computing resources and do not require training samples. Moreover, they are easy to implement on platforms with limited resources and hold great potential for practical implementations The model-driven approaches can be classified into three categories: morphological filtering-drive, human visual system-driven and sparse representation-driven algorithms.
In infrared small target detection, the local structure refers to specific features of the target and its surroundings:
Target Morphology: Small targets (e.g., aircraft, vehicles, or people) appear as bright spots with distinct shapes and sizes.
Edge Information: The boundary between the target and the background is clear and has high contrast, making it a key local feature.
Texture Features: Even in low-resolution images, texture differences provide useful information. For example, the sky and flat ground have uniform textures, while vegetation and buildings show more complex texture.
References
Y. Li, Z. Li, J. Li, J. Yang, A. Siddique, "Robust small infrared target detection using weighted adaptive ring top-hat transformation", Signal Processing, Volume 217, April 2024.