Offshore Platform Extraction Using RadarSat-2 SAR Imagery: A Two-Parameter CFAR Method Based on Maximum Entropy

Area of Science:

• Geospatial analysis
• Remote sensing technology
• Information theory applications

Background:

• Accurate offshore platform identification is vital for oil spill monitoring and resource development.
• Traditional two-parameter constant false alarm rate (CFAR) methods require manual threshold adjustments, impacting efficiency and objectivity.
• Existing methods struggle with accurate target extraction due to subjective parameter settings.

Purpose of the Study:

• To propose an automated and objective two-parameter CFAR target detection method for offshore platforms.
• To leverage information entropy theory and the maximum entropy principle for optimal threshold determination.
• To enhance the accuracy and reliability of offshore platform extraction in remote sensing data.

Main Methods:

• Utilizing a two-parameter CFAR algorithm to generate initial detection thresholds.
• Applying the maximum entropy principle to estimate optimal detection thresholds.
• Employing neighborhood analysis to eliminate false positives, such as ships.

Main Results:

• The proposed method achieved a 97.5% accuracy rate in offshore platform extraction experiments.
• Demonstrated the ability to objectively determine optimal target detection threshold parameters.
• Successfully obtained ideal offshore platform distribution information from South China Sea data.

Conclusions:

• The novel method significantly reduces subjective influence in parameter setting for offshore platform detection.
• It effectively extracts offshore platform targets with high accuracy and reliability.
• This approach offers an objective and efficient solution for offshore platform monitoring and development.