An Improved Detection Method for Hyperspectral Imagery Based on White Gaussian Noise

Area of Science:

• Remote Sensing
• Signal Processing
• Hyperspectral Imaging

Background:

• The adaptive coherence estimator (ACE) is a common hyperspectral target detection method.
• Current ACE methods suffer from low detection efficiency due to deviations from theoretical assumptions.

Purpose of the Study:

• To improve the detection efficiency of the ACE algorithm for hyperspectral images.
• To develop a more universal and adaptive hyperspectral detection method.

Main Methods:

• An improved detection method, termed ACE-WGN, is proposed.
• The method adaptively sets an optimal signal-to-noise (SNR) parameter using spectral angle mapping (SAM).
• White Gaussian noise (WGN) is generated based on the SNR and added to the hyperspectral image data.

Main Results:

• The addition of WGN makes the image data more consistent with ACE's theoretical assumptions.
• The proposed ACE-WGN method significantly improves target detection performance.
• The adaptive SNR parameter setting enhances the method's universality across different hyperspectral images.

Conclusions:

• The ACE-WGN method offers a significant improvement in hyperspectral target detection efficiency.
• The approach is robust and adaptable to various hyperspectral imaging scenarios.
• This method provides a more reliable tool for analyzing hyperspectral data.