Micro 4D Imaging Sensor Using Snapshot Narrowband Imaging Method.

Area of Science:

• Computer Vision
• Optical Imaging
• Sensor Technology

Background:

• Spectral and depth (SAD) imaging is crucial in computer vision.
• Accurate depth estimation and spectral imaging from a single image remain challenging without increasing sensor volume.
• Existing SAD methods often face limitations in complexity and cost.

Purpose of the Study:

• To develop a novel imaging method for accurate depth estimation and spectral capture from a single image.
• To introduce a compact and cost-effective sensor solution for SAD imaging.
• To address the limitations of current SAD imaging techniques.

Main Methods:

• Proposed a micro 4D imaging (M4DI) sensor integrating a monochromatic sensor with a narrow-band spectral filter mosaic.
• Developed a simple remapping algorithm to separate raw images into four narrow spectral bands.
• Implemented a depth estimation algorithm to generate dense 3D depth maps.

Main Results:

• The M4DI sensor maintains the size of a standard monochromatic sensor.
• Achieved quantitative depth estimation with a relative error of less than 7% for objects between 1.1 to 2.8 m.
• Demonstrated proof-of-principle applications in object depth sensing and trajectory tracking.

Conclusions:

• The M4DI sensor offers a simple, low-cost, and computationally efficient solution for SAD imaging.
• This technology advances single-image depth estimation and spectral analysis capabilities.
• The M4DI sensor shows significant potential for various computer vision applications requiring 3D information.