Infrared Small Target Detection via Modified Fast Saliency and Weighted Guided Image Filtering.

Area of Science:

• Computer Vision
• Image Processing
• Remote Sensing

Background:

• Robust small target detection is critical for infrared (IR) search and tracking.
• Existing state-of-the-art (SOTA) methods struggle with edge suppression in complex backgrounds.

Purpose of the Study:

• To develop an effective IR small target detection algorithm.
• To improve background clutter suppression and target detection accuracy.

Main Methods:

• Calculated fast saliency map modulated by steering kernel (SK).
• Utilized weighted guided image filter (WGIF) for edge-preserving smoothing.
• Employed fuzzy sets for background image prediction.
• Computed differential image for target detection via thresholding.

Main Results:

• Effectively suppressed background clutter under strong edge interference.
• Accurately detected small targets with a low false alarm rate.
• Demonstrated performance on four experimental sequences.

Conclusions:

• The proposed method offers a robust solution for IR small target detection.
• It outperforms SOTA methods in complex background scenarios.
• Achieves high accuracy and low false alarms in challenging IR imaging conditions.