Target detection enhancement method through multi-dimensional information differential projection and fusion

Area of Science:

• Optics and Photonics
• Computer Vision and Image Processing
• Remote Sensing

Background:

• Spectral polarization imaging (SPI) offers multi-dimensional data for target detection.
• Challenges remain in SPI for complex scenes: low contrast, structural loss, and clutter.
• Deep learning integration shows promise but requires large datasets and processing.

Purpose of the Study:

• To develop an innovative target detection enhancement method for SPI systems.
• To improve the detection of targets in complex natural environments.
• To enhance scenario comprehension and detection efficiency without extensive computation.

Main Methods:

• Developed a novel target retrieval algorithm based on angle of polarization purification.
• Implemented a multi-dimensional information differential projection and fusion strategy.
• Utilized self-developed SPI systems for data acquisition and processing.

Main Results:

• The proposed method effectively exploits and enhances multi-dimensional distinctions between targets and backgrounds.
• Fused images successfully highlight targets with shielding, camouflage, and shape loss.
• Demonstrated efficacy in complex environments like grasslands, woodlands, and sky.

Conclusions:

• The innovative SPI method significantly improves target detection in challenging scenarios.
• The approach enhances detection efficiency and situational awareness for applications like military reconnaissance.
• The method offers advantages due to reduced computational requirements and no need for model training.