Frequency-modulated continuous-wave LiDAR compressive depth-mapping.

Area of Science:

• Optics and Photonics
• Computer Vision
• Signal Processing

Background:

• Traditional LiDAR systems often rely on raster scanning, which can be time-consuming.
• Achieving high signal-to-noise ratios in depth mapping can be challenging with detector-array based schemes.

Purpose of the Study:

• To develop an inexpensive architecture for converting frequency-modulated continuous-wave (FMCW) LiDAR into a compressive-sensing (CS) based depth-mapping camera.
• To reduce the number of measurements required for depth map acquisition using compressive sensing techniques.

Main Methods:

• Utilizing compressive sensing instead of raster scanning to acquire depth information.
• Employing a system ideally requiring two difference detectors, leveraging heterodyne detection and background subtraction.
• Solving two linear equations with convex optimization methods to extract depth data from compressed measurements.

Main Results:

• Significantly reduced number of measurements compared to traditional scanning methods.
• Achieved higher signal-to-noise ratios over detector-array based schemes.
• Enabled faster scene scanning and efficient data storage (2m points from an n-pixel scene, m < n).

Conclusions:

• The proposed architecture offers an inexpensive and efficient solution for depth mapping using FMCW LiDAR.
• Compressive sensing enables faster data acquisition and improved signal quality in depth-mapping applications.
• The method allows for straightforward depth extraction through convex optimization, making it practical for real-world implementation.