Efficient adaptation of complex-valued noiselet sensing matrices for compressed single-pixel imaging.

Area of Science:

• Optics and photonics
• Signal processing
• Computational imaging

Background:

• Compressed-sensing single-pixel detectors rely on measurement matrices with minimal mutual coherence, often using discrete noiselets and Haar wavelets.
• Existing methods face limitations in efficiently utilizing binary spatial light modulators with complex-valued sampling functions.

Purpose of the Study:

• To propose an efficient method for object sampling in single-pixel cameras using complex-valued, nonbinary noiselet functions.
• To enable the determination of complex noiselet coefficients from a minimal number of binary measurements.
• To develop a computationally efficient algorithm for real-time pattern generation.

Main Methods:

• Utilizing complex-valued and nonbinary noiselet functions for object sampling in single-pixel cameras.
• Employing binary spatial light modulators and incoherent illumination.
• Modifying the complex fast noiselet transform for efficient real-time pattern generation using integer calculations.
• Determining m complex noiselet coefficients from m+1 binary sampling measurements.

Main Results:

• Successfully determined complex noiselet coefficients from binary sampling measurements.
• Introduced a computationally efficient modification to the complex fast noiselet transform.
• Demonstrated real-time generation of binary noiselet-based patterns.
• Experimentally verified the proposed method with a single-pixel camera system.

Conclusions:

• The proposed method offers an efficient approach for object sampling in single-pixel cameras.
• The technique effectively utilizes binary spatial light modulators for complex noiselet coefficient determination.
• The computationally efficient algorithm facilitates real-time applications in compressed-sensing imaging.