Sharpness evaluation algorithm for nailfold microvascular images

An Huang, Zhiwei Wu, Hao Yin, Qianyao Ye, Junzhao Liang, Jianan Lin, Mugui Xie, Cong Ye, Xiaosong Li, Yanxiong Wu

Abstract: Observing and analyzing nailfold microvascular images are essential for evaluating human health, but effective analysis requires high-clarity images. Autofocus technology enables the efficient and convenient acquisition of high-quality images for clinical examinations. However, successful autofocus depends on the accuracy of the sharpness evaluation algorithm. Owing to the low contrast of nailfold microvascular images and the inevitable dithering artifacts and stray light produced during acquisition, existing algorithms cannot accurately evaluate sharpness, leading to autofocus failure. To address this problem, this study proposes a sharpness evaluation algorithm that uses histogram equalization to enhance the high-frequency information of the images, applies wavelet transform to reduce the influence of dithering artifacts and stray light, and calculates the edge gradient to evaluate sharpness quantitatively. Five mainstream evaluation algorithms were compared to validate the performance of the proposed algorithm, and the results show that it outperforms the others and accurately determines the clearest image. Compared with other algorithms, the clearest image exhibits the best visual effect, with a narrow width of the evaluation curve that is 27–37% higher. The algorithm proposed in this article has better real-time applicability.