Extremely Robust Remote-Target Detection Based on Carbon Dioxide-Double Spikes in Midwave Spectral Imaging.

Area of Science:

• Remote sensing
• Infrared spectroscopy
• Optical surveillance

Background:

• Ship-target detection from coast is challenging due to background clutter like clouds and sea glint.
• Conventional methods use spatial or temporal information, often leading to false positives.
• Existing infrared surveillance struggles with distinguishing targets from environmental interference.

Purpose of the Study:

• To propose a novel infrared ship-target detection method using spectral imaging.
• To address the challenge of false positives in sea surveillance.
• To enhance the robustness of ship detection in cluttered marine environments.

Main Methods:

• Developed a carbon dioxide-double spike (CO2-DS) detection technique utilizing midwave spectral imaging.
• Exploited the spectral feature of a broader hot CO2 emission band compared to atmospheric CO2 absorption.
• Employed a directional-mean subtraction filter (D-MSF) for CO2 spike detection and joint analysis for final target identification.

Main Results:

• The CO2-DS detection method demonstrated an extremely low number of false positives caused by background clutter.
• Ship CO2 plume signatures were uniquely identified in the double spikes across different spectral bands.
• Experimental results validated the extreme robustness of the CO2-DS detection in remote sea environments using midwave FTIR.

Conclusions:

• The CO2-DS detection method offers a highly effective solution for infrared ship-target detection.
• This spectral imaging approach significantly improves detection accuracy by minimizing false positives.
• The technique shows extreme robustness, making it suitable for challenging sea surveillance applications.