Enhancing the Image Resolution of Acoustic Microscopy with FISTA

Scanning acoustic microscopy (SAM) systems use ultrasound to acquire the internal acoustic properties of scanned subjects and then depict these subjects in the form of a stack of 2D images. SAM systems are particularly useful because they enable inspection of the internal structures of subjects at a...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2021 5th International Conference on Automation, Control and Robots (ICACR) s. 7 - 11
Hlavní autoři: Wu, Xi, Lu, Ke, Liang, Hao, Xue, Jian
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 25.09.2021
Témata:
Deconvolution
Filtering
FISTA
Image resolution
Microscopy
scanning acoustic microscopy
super-resolution
Ultrasonic imaging
Wavelet transforms
Wiener filters
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Scanning acoustic microscopy (SAM) systems use ultrasound to acquire the internal acoustic properties of scanned subjects and then depict these subjects in the form of a stack of 2D images. SAM systems are particularly useful because they enable inspection of the internal structures of subjects at a microscopic level without causing any damage. However, the unavoidable noise introduced while conducting the scan can cause the acquired images to be blurry, thus limiting the resolution of SAM systems. Unfortunately, traditional filtering and deconvolution methods have only limited abilities to improve the time-domain resolution and show poor stability when the signal-to-noise ratio (SNR) of the image is low. To overcome these disadvantages, we propose a sparse devolution method based on a fast iterative shrinkage-thresholding algorithm (FISTA) to improve the quality of the images obtained. In the experiments, we compared FISTA with the Wiener filter and wavelet transform-based denoising methods. The results show that the FISTA method is better than the other traditional methods in terms of structural similarity, mean square error, peak SNR, and improved SNR.
DOI:10.1109/ICACR53472.2021.9605181