A Method for Medical Microscopic Images' Sharpness Evaluation Based on NSST and Variance by Combining Time and...

Area of Science:

• Microscopy
• Image Processing
• Computational Imaging

Background:

• Microscope autofocus is critical for image quality.
• Existing autofocus methods can be sensitive to noise.
• Accurate sharpness evaluation is essential for reliable autofocus.

Purpose of the Study:

• To develop a robust algorithm for microscopic image sharpness evaluation.
• To improve the noise immunity and accuracy of microscope autofocus.
• To enhance the performance of autofocus systems in challenging imaging conditions.

Main Methods:

• Image decomposition using non-subsampled shearlet wave transform (NSST).
• Variance calculation on NSST sub-band images to determine coefficient energy.
• Sharpness evaluation based on the ratio of high- to low-frequency sub-band energy.

Main Results:

• The proposed algorithm demonstrates superior noise immunity compared to existing methods.
• High sensitivity in sharpness evaluation is maintained.
• Experimental results validate the algorithm's effectiveness.

Conclusions:

• The NSST-based variance algorithm offers a significant improvement for microscope autofocus.
• The method provides a reliable approach for sharpness evaluation in noisy microscopic images.
• This technique enhances the accuracy and robustness of automated focusing in microscopy.