Restoration of turbulence-degraded extended object using the stochastic parallel gradient descent algorithm:...

Area of Science:

• Optics and Photonics
• Image Processing
• Computational Science

Background:

• Turbulence degrades image quality in optical systems.
• Adaptive optics (AO) systems are crucial for image restoration.
• Developing efficient image performance metrics is essential for AO system control.

Purpose of the Study:

• To simulate an AO system for restoring turbulence-degraded extended objects.
• To identify a practical image performance metric for the SPGD algorithm using only gray-level information.
• To evaluate the AO system's restoration capability under varying turbulence conditions.

Main Methods:

• Simulation of an adaptive optics system incorporating a 61-element deformable mirror.
• Application of the Stochastic Parallel Gradient Descent (SPGD) algorithm to optimize deformable mirror actuator voltages.
• Evaluation of image performance metrics, focusing on gray level variance, for SPGD control.

Main Results:

• The gray level variance function demonstrated superior performance as an image quality metric compared to gradient-based metrics.
• The simulated AO system successfully restored images degraded by various turbulence strengths.
• The chosen performance metric proved effective for the SPGD algorithm in image restoration.

Conclusions:

• The gray level variance is a viable and effective image performance metric for AO systems controlled by SPGD.
• The simulated AO system effectively compensates for turbulence-induced aberrations.
• This approach offers a practical method for enhancing image quality in challenging optical environments.