Infrared Small Target Detection Based on Weighted Local Coefficient of Variation Measure

Area of Science:

• Computer Science
• Electrical Engineering
• Image Processing

Background:

• Infrared (IR) small target detection is crucial for infrared search and track (IRST) systems.
• Complex backgrounds pose significant challenges, leading to high false alarm rates and missed detections in current algorithms.

Purpose of the Study:

• To propose a novel and robust method for IR small target detection.
• To address the limitations of existing algorithms in complex background scenarios.

Main Methods:

• A three-stage approach involving preprocessing, a detection stage with background suppression module (BSM) and local coefficient of variation (LCV) module, and adaptive threshold segmentation.
• BSM utilizes a three-layer window considering target anisotropy and grayscale distribution.
• LCV leverages discrete statistical properties of target grayscale.

Main Results:

• The proposed weighted local coefficient of variation (WLCV) method significantly enhances small targets and suppresses backgrounds.
• Experimental results demonstrate superior robustness across different target sizes and background types compared to existing methods.
• The WLCV method achieves higher detection accuracy.

Conclusions:

• The WLCV method offers a robust solution for IR small target detection in challenging environments.
• This technique improves the performance of IRST systems by increasing detection accuracy and reducing false alarms.