Evaluating denoising filters in a real-world situation is hard since information about noises (level, type, distribution, etc.) is unknown. Hence, determining the performance of denoising filters is a challenge. To solve this problem, three groups of images are synthesized to evaluate the performance of denoising filters:
(1) spatial resolution test images: this group is useful to evaluate the performance of filters in preserving small details of objects (Figures A-B);
(2) Contrast resolution test images: this group is used to evaluate the performance of filters
in preserving objects with low contrast resolution (Figures C-D);
(3) ultrasound simulation images: these images were synthesized to mimic the ideal (noisefree) real ultrasound images (Figures E-F).
If you use this dataset, please cite:
Ardakani, AA, Mohammadi, A, Faeghi, F, Acharya, UR. Performance evaluation of 67 denoising filters in ultrasound images: A systematic comparison analysis. Int J Imaging Syst Technol. 2023; 33( 2): 445- 464. doi:10.1002/ima.22843
Synthesized test (A–D) and ultrasound (E–G) images for denoising filters evaluation.
