Novel Sparse Recovery Algorithms for 3D Debris Localization using Rotating Point Spread Function Imagery

Chao Wang, Robert J. Plemmons, Sudhakar Prasad, Raymond H. Chan, Mila Nikolova

Abstract

An optical imager that exploits off-center image rotation to encode both the lateral and depth coordinates of point sources in a single snapshot can perform 3D localization and tracking of space debris. When actively illuminated, unresolved space debris, which can be regarded as a swarm of point sources, can scatter a fraction of laser irradiance back into the imaging sensor. Determining the source locations and fluxes is a large-scale sparse 3D inverse problem, for which we have developed efficient and effective algorithms based on sparse recovery using non-convex optimization. Numerical simulations illustrate the efficiency and stability of the algorithms.

Introduction

As Earth's orbit becomes increasingly crowded, tracking "microdebris" (sub-centimeter sized objects) is critical for satellite safety. While radar is excellent for long-range detection, optical telescopes offer much higher precision at closer ranges (hundreds of meters to kilometers). However, standard optical imaging has a major weakness: it is effective at determining lateral $(x, y)$ coordinates but struggles with depth $(z)$ due to a limited field of depth. Traditional solutions require multiple cameras (parallax) or complex laser ranging. This paper addresses the challenge of 3D localization from a single snapshot by using a computational imaging approach.

Methods

The authors employ a "Rotating PSF" (Point Spread Function) technique. By placing a specific annular phase mask on the telescope's aperture, the image of a point source (like a piece of debris) rotates predictably as it moves out of focus. This rotation angle encodes the depth information directly into the 2D snapshot.

To reconstruct 3D coordinates from these images, the task is framed as a large-scale sparse 3D inverse problem. The core contributions include:

  1. Non-Convex Regularization: Instead of standard $\ell_1$ (Lasso) regularization, which can bias results, the authors use non-convex penalties like CEL0 (for Gaussian noise) and a custom NC penalty (for Poisson noise). These better approximate the ideal $\ell_0$ norm.
  2. IRL1 Algorithm: An Iteratively Reweighted $\ell_1$ algorithm is used to solve these non-convex problems efficiently.
  3. Novel Flux Estimation: For Poisson noise, a new fixed-point iterative scheme accurately calculates debris brightness (flux).
  4. Post-Processing: A centroid-based clustering method handles "over-fitting" where single objects might be identified as clusters.
Non-convex penalty functions used for sparse recovery: (a) CEL0 penalty for Gaussian noise and (b) NC penalty for Poisson noise.

The study utilizes simulated datasets (96x96 pixels) representing swarms of 5 to 40 point sources in 3D space, subjected to Gaussian (read-out) and Poisson (photon) noise models.

Results and Discussion

The proposed algorithms consistently outperformed standard $\ell_2$-$\ell_1$ methods. In Gaussian noise tests with 5–15 sources, the CEL0 algorithm showed a >10% improvement in precision. Even in high-density scenarios (40 sources), the non-convex approach maintained a precision of ~53% compared to 42% for standard methods.

For Poisson noise, the new iterative flux estimation significantly improved accuracy, keeping relative errors mostly within the [0, 0.1] range, whereas $\ell_1$ methods often exceeded 0.3. The algorithms also maintained high recall, identifying nearly all true sources even in noisy conditions. These results highlight that non-convex penalties provide a "clearer" selection of true sources by avoiding the signal "shrinkage" typical of $\ell_1$ regularization.

Conclusion

This work demonstrates that 3D localization of space debris can be achieved with high precision from a single snapshot using rotating PSFs and non-convex optimization. While currently validated through simulations, future work will involve testing on real data with fabricated phase masks. The authors also aim to extend this approach to snapshot multi-spectral imaging to identify debris material composition alongside 3D localization.

Summary updated

Abstract

An optical imager that exploits off-center image rotation to encode both the lateral and depth coordinates of point sources in a single snapshot can perform 3D localization and tracking of space debris. When actively illuminated, unresolved space debris, which can be regarded as a swarm of point sources, can scatter a fraction of laser irradiance back into the imaging sensor. Determining the source locations and fluxes is a large-scale sparse 3D inverse problem, for which we have developed efficient and effective algorithms based on sparse recovery using non-convex optimization. Numerical simulations illustrate the efficiency and stability of the algorithms.
BibTeX
@inproceedings{wang2018novel,
  author = {Chao Wang and Robert J. Plemmons and Sudhakar Prasad and Raymond H. Chan and Mila Nikolova},
  title = {{Novel Sparse Recovery Algorithms for 3D Debris Localization using Rotating Point Spread Function Imagery}},
  booktitle = {The Advanced Maui Optical and Space Surveillance Technologies Conference},
  year = {2018},
}
Figure