The Alignment of High-resolution Solar Prominence Images Observed by the New Vacuum Solar Telescope

TL;DR

This paper introduces a new high-precision alignment algorithm for solar prominence images captured by the NVST, addressing challenges of low contrast and structural complexity to improve accuracy and robustness.

Contribution

The study develops a comprehensive alignment method that combines optical flow, edge gradients, and cross-correlation tailored for prominence images, surpassing existing approaches.

Findings

The proposed method achieves higher alignment accuracy than traditional techniques.

It demonstrates robustness across diverse prominence postures.

The algorithm effectively handles low-contrast and complex prominence structures.

Abstract

High spatial resolution observation of solar prominence is an important observation subject of the New Vacuum Solar Telescope (NVST). While the current level of observation and image reconstruction technologies for solar prominences are advanced, a significant challenge remains in achieving high-precision alignment among high-resolution prominence images observed at different times and different off-bands. Existing alignment approaches either become ineffective or yield low accuracy, and always require manual intervention during the alignment. These limitations are largely due to the stronger edge gradient and lower structural contrast of the prominence images compared with the solar surface ones. In response to this challenge, our study aims to develop an effective and robust algorithm for high-precision alignment of the prominence images. We thoroughly consider the unique structural characteristics of prominence images and the specific application conditions of various alignment algorithms. Consequently, we propose a comprehensive alignment method that incorporates the optical flow field of the solar surface, the gradient of the solar edge, and the cross-correlation within the solar prominence region. This method is designed to accurately determine the movement displacements among the prominence images. Our results demonstrate that this alignment method excels in both accuracy and robustness, making it well-suited for handling the diverse postures of solar prominence images captured by the NVST.