Aircraft detection from VHR images based on circle-frequency filter and multilevel features

Area of Science:

• Remote Sensing
• Computer Vision
• Image Analysis

Background:

• Aircraft detection in very high-resolution (VHR) imagery is crucial for various applications.
• Existing methods may face challenges in accurately distinguishing aircraft from complex backgrounds in VHR images.

Purpose of the Study:

• To propose a novel and effective detector for automatic aircraft detection in VHR remote sensing images.
• To improve the accuracy and robustness of aircraft identification in challenging imaging conditions.

Main Methods:

• Utilized a circle-frequency filter (CF-filter) for efficient extraction of candidate aircraft locations from large VHR images.
• Developed a multi-level feature model incorporating Robust Hue Descriptor and Histogram of Oriented Gradients to represent local appearance and spatial layout.
• Integrated these components to create a comprehensive aircraft detection system.

Main Results:

• The proposed detector demonstrated superior performance in accurately distinguishing aircraft from background clutter.
• Experimental results validated the effectiveness of the CF-filter and the multi-level feature model for aircraft detection.
• The method showed high accuracy in identifying aircraft in diverse VHR remote sensing scenarios.

Conclusions:

• The novel detector offers a significant advancement in automatic aircraft detection from VHR images.
• The combination of CF-filter and the multi-level feature model provides a robust solution for remote sensing applications.
• This research contributes to improved capabilities in aerial surveillance and monitoring through enhanced image analysis.