New strategy for high-dimensional single-pixel imaging.

Area of Science:

• Optics and Photonics
• Computational Imaging
• Computer Vision

Background:

• Single-pixel imaging (SPI) offers a cost-effective alternative to traditional imaging systems by minimizing detector resolution requirements.
• Existing SPI techniques are primarily limited to 2D imaging or basic 3D reconstruction, lacking comprehensive high-dimensional data acquisition capabilities.
• There is a need for advanced SPI methods that can simultaneously capture depth, color, and texture information with improved accuracy and reduced computational complexity.

Purpose of the Study:

• To propose and demonstrate a comprehensive single-pixel imaging strategy for high-dimensional data acquisition, including 3D shape, color texture, and viewpoint expansion.
• To develop a low-cost 3D imaging system using SPI, incorporating structured illumination and grating-based height encoding.
• To enhance the accuracy of 3D shape fusion and color fidelity through novel data processing techniques.

Main Methods:

• A low-cost single-pixel 3D imaging system employing a raster scanner for structured illumination and a grating for height encoding.
• Utilization of two single-pixel detectors (SPDs) from different angles to capture height information, with a modified total variation criterion for error reduction in shape fusion.
• Integration of a third SPD with bandpass filters for color texture acquisition and a viewpoint switching method inspired by shape-from-shading for improved color fidelity.

Main Results:

• Successful implementation of a single-pixel 3D imaging scheme with reduced blocking area and enhanced shape fusion accuracy.
• Acquisition of color texture information using multiple SPDs and bandpass filters, demonstrating capability for higher dimensional data capture.
• Improved color fidelity through a novel viewpoint switching method, showcasing the potential for comprehensive high-dimensional imaging with SPI.

Conclusions:

• The proposed SPI strategy represents a significant advancement in high-dimensional imaging, offering a versatile and cost-effective solution.
• This work provides a foundational demonstration for the application of SPI in the acquisition, reconstruction, and fusion of complex, high-dimensional image data.
• The developed technique has broad implications for various fields requiring detailed 3D and multi-spectral imaging with minimal hardware complexity.