High-resolution remotely sensed small target detection by imitating fly visual perception mechanism

Area of Science:

• Remote Sensing
• Computer Vision
• Biomimicry

Background:

• Small target detection in high-resolution remote sensing data presents significant challenges.
• Existing methods struggle with complex natural environments.
• The visual system of flies offers a potential model for advanced information processing.

Purpose of the Study:

• To develop a novel small target detection model inspired by the fly visual system.
• To enhance the speed and accuracy of small target detection in remote sensing.
• To provide a theoretical framework for fly-inspired visual perception in remote sensing.

Main Methods:

• Constructed a characterized model of information perception based on fly visual system theory.
• Developed a fly-imitated visual system for information processing in remote sensing.
• Designed a small target detector and algorithm simulating fly visual mechanisms (acquisition, compression, fusion, pool cell function, nonlinear self-adaptation).

Main Results:

• Experimental verification of the proposed small target detection model's feasibility.
• Demonstrated the rationality of the fly-imitated visual perception method.
• Achieved fast and accurate small target detection under complex environmental conditions.

Conclusions:

• The fly-imitated visual perception method offers a promising approach for small target detection in high-resolution remote sensing.
• The developed model provides a robust theoretical basis for future research in this area.
• This biomimetic strategy effectively addresses limitations of current detection techniques.