• Homepage
  • European Microscopy Congress 2020 Abstract Database
  • Wavelet_Denoise Fiji/ImageJ plugin for filtering/denoising microscopic images

    • Wavelet_Denoise Fiji/ImageJ plugin for filtering/denoising microscopic images

    Abstract number
    1443
    Event
    European Microscopy Congress 2020
    DOI
    10.22443/rms.emc2020.1443
    Corresponding Email
    [email protected]
    Session
    DHA.3 - Machine assisted acquisition and analysis of microscopy data
    Authors
    Dr. Martin Čapek (2), Dr. Ivan Novotný (2), Dr. Michaela Efenberková (2), Dr. Helena Chmelová (2), Dr. Barbora Radochová (1), Dr. Jiří Janáček (1), Dr. Ondrej Horvath (2)
    Affiliations
    1. Department of Biomathematics, Institute of Physiology of the CAS
    2. Light Microscopy Core Facility, Institute of Molecular Genetics of the CAS
    Keywords

    Fiji, plugin, wavelet transform, image filtration

    Abstract text

    Filtering microscopic images is difficult. If we try to remove undesired image parts, e.g., noise, inhomogenous background, reconstruction artifacts, by standard filtration techniques, we may lose resulting resolution that was acquired by hi-tech microscopy systems. 

    Standard filtration methods include convolution-based techniques, like Gaussian smoothing, or Fourier-based techniques [1]. These techniques suppose that noise and artifacts are small image parts that can be removed by some kind of convolution or by lowering the range of image frequencies in case of Fourier filtering. These techniques work nicely, however, they do blur and may increase the size of structures, which is not welcome. Thus, another technique of image filtration that is able to preserve the size of structures and original resolution of microscopic data as much as possible would be desired.

    One of promising and relatively modern techniques applied in digital image processing is a discrete wavelet transform (DWT) [1]. DWT is able to distinguish, that one object is large and relatively intensity homogenous and other objects in the same picture are subtle with sudden intensity changes, which is advantageous for filtering.

    There have been developed several tools for wavelet transformation of images till now, e.g., web-based demos [2], Matlab (Wavelet Toolbox) and its applications [3-6], stand-alone applications [7-9] and plugins [8-12] for ImageJ/Fiji [13-14].

    Mostly, the above approaches are demos or simple applications for DWT. The exceptions are tools in Matlab, but it is not free. Other exceptions, both for ImageJ/Fiji, are Xlib library [10] that offers DWT of images with various wavelet families, but without filtering wavelet coefficients, and Fractional Wavelet Module plugin [11] with various filter coefficients possibilities prior image synthesis, however, which is not offering some of the most common wavelet families used nowadays.

    Thus, in our opinion, there is still missing a practical, freely available tool that does interactively DWT based filtering of large 2D/3D microscopic images using modern wavelet families.

    Therefore, we implemented a Fiji/ImageJ plugin called Wavelet_Denoise, see Fig. 1. It demonstrates main possibilities of applying DWT to images: (i) Decomposition of an input picture using various wavelet filters and levels of details with decomposed image visualization; (ii) Effects of back transformation of the picture with some level of suppression or denoising (iii) of wavelet coefficients.

    Advantages of Fiji are that the package reads images using a plethora of various microscopic formats; users can easily install the plugin through a menu command and the plugin can support processing 3D images in Z-stacks.

    In this work we also demonstrate the application of the plugin and wavelet filtering for removal of reconstruction artifacts and undesirable background in images acquired by super-resolved structured illumination microscopy.

    References

    [1] RC Gonzales, RE Woods (2008). Digital image processing. 3rd ed. Pearson Prentice Hall, 954 p.

    [2] D Sage, M Unser (2003). Teaching Image-Processing Programming in Java. IEEE Signal Processing Magazine, 20(6):43–52.

    [3] F Luisier, T Blu (2008). SURE-LET Multichannel Image Denoising: Interscale Orthonormal Wavelet Thresholding. IEEE Transactions on Image Processing, 17(4): 482–492.

    [4] F Luisier, T Blu, M Unser (2007). A New SURE Approach to Image Denoising: Interscale Orthonormal Wavelet Thresholding. IEEE Transactions on Image Processing, 16(3): 593–606.

    [5] WavBox Software Library, toolsmiths.com

    [6] WaveLab, statweb.stanford.edu/~wavelab

    [7] EG Roselló, JG Dacosta, MJ Lado, AJ Méndez, J Sampedro, MP Cota (2014). Visual Wavelet-Lab: An object-oriented library and a GUI application for the study of the wavelet transform. Computer Applications in Engineering Education. 22(1): 23–32.

    [8] codeproject.com/Articles/20869/D-Fast-Wavelet-Transform-Library-for-Image-Pro-ces

    [9] accord-framework.net/samples.html

    [10] imagej.net/Xlib#Wavelets_2D

    [11] M Unser, T Blu (2000). Fractional Splines and Wavelets. SIAM Review. 42(1): 43–67.

    [12] imagej.nih.gov/ij/plugins/haar-wavelet-filter.html

    [13] imagej.nih.gov

    [14] Fiji.sc

    [15] Supported by the MEYS CR: LM2015062, CZ.02.1.01/0.0/0.0/16_013/0001775, LO1419. 

    Return to listing