[Paper Review] Fast localization and single-pixel imaging of the moving object using time-division multiplexing

Zijun Guo, Wenwen Meng+6 more|arXiv (Cornell University)|Aug 15, 2022

Random lasers and scattering media4 citations

TL;DR

This paper proposes a time-division multiplexing method that decouples motion and image information in single-pixel imaging to enable anti-motion blur reconstruction of fast-moving objects. By alternately encoding geometric moment patterns (for position) and Hadamard patterns (for image), the method achieves 5.55 kHz localization frequency and pseudo 512×512 resolution imaging of objects rotating at up to 0.5 rad/s.

ABSTRACT

When imaging moving objects, single-pixel imaging produces motion blur. This paper proposes a new single-pixel imaging method, which can achieve anti-motion blur imaging of a fast-moving object. The geometric moment patterns and Hadamard patterns are used to alternately encode the position information and the image information of the object with time-division multiplexing. In the reconstruction process, the object position information is extracted independently and combining motion-compensation reconstruction algorithm to decouple the object motion from image information. As a result, the anti-motion blur image and the high frame rate object positions are obtained. Experimental results show that for a moving object with an angular velocity of up to 0.5rad/s relative to the imaging system, the proposed method achieves a localization frequency of 5.55kHz, and gradually reconstructs a clear image of the fast-moving object with a pseudo resolution of 512x512. The method has application prospects in single-pixel imaging of the fast-moving object.

Motivation & Objective

• Address motion blur in single-pixel imaging of fast-moving objects.
• Decouple object motion information from image data to enable motion-compensated reconstruction.
• Achieve high-speed localization and clear image reconstruction for rapidly moving targets.
• Enable real-time imaging of dynamic scenes with minimal motion artifacts using a single-pixel detector.

Proposed method

• Use geometric moment patterns to encode and extract object position information via time-division multiplexing.
• Alternate between geometric moment patterns and Hadamard patterns in temporal encoding to separate motion and image data.
• Apply a motion-compensation reconstruction algorithm that leverages independently extracted position data to correct image distortions.
• Reconstruct the image using compressed sensing principles with motion-corrected projections.
• Utilize time-division multiplexing to allocate distinct temporal slots for position and intensity encoding.
• Leverage the sparsity of motion in angular velocity to enhance localization accuracy and frame rate.

Experimental results

Research questions

• RQ1 Can motion blur in single-pixel imaging be effectively mitigated for fast-moving objects?
• RQ2 Can object position be estimated independently and with high temporal resolution using time-division multiplexing?
• RQ3 Can the decoupling of motion and image information enable high-frame-rate localization and clear image reconstruction?
• RQ4 What is the maximum angular velocity for which motion-compensated single-pixel imaging remains effective?
• RQ5 Can the method achieve sub-millisecond localization frequency with minimal image degradation?

Key findings

• The method achieves a localization frequency of 5.55 kHz for moving objects.
• For objects with an angular velocity of up to 0.5 rad/s, the system successfully reconstructs clear images with a pseudo-resolution of 512×512.
• Motion blur is effectively suppressed through independent extraction and compensation of object position data.
• The time-division multiplexing of geometric moment and Hadamard patterns enables simultaneous high-speed localization and image reconstruction.
• The motion-compensation algorithm successfully decouples motion from image information, improving reconstruction fidelity.
• The approach demonstrates practical feasibility for single-pixel imaging of fast-moving targets in dynamic environments.