rus
Issue #3-4/2024
A.I.Akhmetova, D.I.Yaminsky, I.V.Yaminsky
FEMTOSCAN ONLINE: IMAGE PROCESSING AND FILTERING
FEMTOSCAN ONLINE: IMAGE PROCESSING AND FILTERING
DOI: https://doi.org/10.22184/1993-8578.2024.17.3-4.178.183
The unique capabilities of the atomic force microscope (AFM), including super-high-resolution imaging, nanomanipulation, and the ability to operate under physiological conditions, have opened up exciting research opportunities in biology and biomedicine. AFM imaging has helped reveal the fine structure of bacterial cell walls at the nanoscale and how they are altered by antimicrobial treatment. This paper discusses the functions of FemtoScan Online software in relation to processing probe microscopy images, improving image quality reducing noise and increasing the informativeness of data.
The unique capabilities of the atomic force microscope (AFM), including super-high-resolution imaging, nanomanipulation, and the ability to operate under physiological conditions, have opened up exciting research opportunities in biology and biomedicine. AFM imaging has helped reveal the fine structure of bacterial cell walls at the nanoscale and how they are altered by antimicrobial treatment. This paper discusses the functions of FemtoScan Online software in relation to processing probe microscopy images, improving image quality reducing noise and increasing the informativeness of data.
INTRODUCTION
In probe microscopy, experimental data plays a decisive role in obtaining new information about the structure of matter. It is often necessary to work with samples with complex topography, and the features need to be characterised, counted, and digitised. For this purpose it is necessary to use numerous functions. FemtoScan Online software copes with this task [1]. In addition to the basic functions of averaging and subtraction of large relief, which we described earlier [2], it is possible to use more complex and more delicate operations that allow to transform the obtained image.
RESEARCH METHODS
Fourier filtering. When subtracting the background we have to face a typical problem in physics: the background and the signal (relief of the microstructure of interest) are of the same order. In this case, an alternative method of macrorelief removal – Fourier filtering will help.
The idea is simple – macro-relief corresponds to trigonometric functions with low frequencies, so if we remove the low-frequency component (a small fragment near the origin, Fig.1) from the Fourier image, these image components will be filtered out: all the micro-relief will remain and the macro-relief will disappear.
In order to remove the noise component without reducing the sharpness of the microstructure display, we can also use Fourier filtering. Having constructed the Fourier image, we see (Fig.2) that in the high frequency region – further from the origin – there are several bright reflexes. These are the reflex bands along the X axis and a few bright reflexes close to the Y axis. They should be removed, then after the inverse Fourier transform the noise in the image will be much less. You can do more radical and remove all the high frequencies in the image, as shown in Fig.3. The result of the inverse transformation is shown in Fig.4.
You can change the contrast of the Fourier image: to do this, select one of the three options of the Contrast command from the Fourier menu: More (increases contrast), Less (decreases it) or Restore (restores the original setting). If you have selected some area of the spectrum, you can zero its inner or outer part. To do this, there are commands Zero Inside and Zero Outside, available by right-clicking and also from the Fourier menu. Similarly, you can increase or decrease the contribution of selected frequencies using the Increase or Decrease commands.
By amplifying, attenuating and zeroing the selected frequencies, the changes in the Fourier spectrum are immediately reflected in the surface image. Weakening of the low-frequency component in the image results in removal of macro relief.
Histogram. Histogram is often a good way to quickly estimate the height of film or other uniform objects. After executing the Histogram command, a window containing distribution of surface image points by heights appears. When moving dashed vertical lines in the bottom line of the window, numerical data are displayed (Fig.5):
Х0 – is the surface height corresponding to the left mark on the histogram,
Х1 – is the surface height corresponding to the right mark on the histogram,
Х1–Х0 – is the difference in heights between these two levels,
% – percentage of the number of surface points whose height is between the selected heights X0 and X1.
Fig.5 shows the histogram of the image X of potato virus: there is a clear left peak corresponding to the distribution of pure substrate points, the right peak (small) characterises the distribution of particles by height.
Using the Crop command from the Histogram menu, the height of each point outside the vertical lines becomes equal to the height indicated by the corresponding vertical line. These changes are immediately reflected in the surface image. The histogram is also rebuilt again and can have a completely new appearance.
CONCLUSIONS
Atomic force microscopy (AFM) and high-speed scanning have greatly advanced the study of biological objects and structures from single molecules to the cellular level. To facilitate interpretation of limited resolution images, data processing and filtering play an increasingly important role in understanding AFM measurements. When software can be used to characterise the surface topography features of a sample, its structure, and calculate geometric features, it greatly complements the computational analysis of objects.
FemtoScan Online software, along with a user-friendly interface for image processing and filtering, allows to obtain many quantitative features thanks to clever algorithms and high-quality visualisation, which certainly contributes to the understanding of processes beyond topographic images [3]. This work illustrates the capabilities of FemtoScan Online software and emphasises the importance of AFM data processing to complement experimental observations.
ACKNOWLEDGMENTS
The work of the staff of the Physical Department was carried out under the state order with the financial support of the Physical Department of the Lomonosov Moscow State University (Registration subject 122091200048-7). FemtoScan Online software is provided by Advanced Technologies Center (www.nanoscopy.ru).
PEER REVIEW INFO
Editorial board thanks the anonymous reviewer(s) for their contribution to the peer review of this work. It is also grateful for their consent to publish papers on the journal’s website and SEL eLibrary eLIBRARY.RU.
Declaration of Competing Interest. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
In probe microscopy, experimental data plays a decisive role in obtaining new information about the structure of matter. It is often necessary to work with samples with complex topography, and the features need to be characterised, counted, and digitised. For this purpose it is necessary to use numerous functions. FemtoScan Online software copes with this task [1]. In addition to the basic functions of averaging and subtraction of large relief, which we described earlier [2], it is possible to use more complex and more delicate operations that allow to transform the obtained image.
RESEARCH METHODS
Fourier filtering. When subtracting the background we have to face a typical problem in physics: the background and the signal (relief of the microstructure of interest) are of the same order. In this case, an alternative method of macrorelief removal – Fourier filtering will help.
The idea is simple – macro-relief corresponds to trigonometric functions with low frequencies, so if we remove the low-frequency component (a small fragment near the origin, Fig.1) from the Fourier image, these image components will be filtered out: all the micro-relief will remain and the macro-relief will disappear.
In order to remove the noise component without reducing the sharpness of the microstructure display, we can also use Fourier filtering. Having constructed the Fourier image, we see (Fig.2) that in the high frequency region – further from the origin – there are several bright reflexes. These are the reflex bands along the X axis and a few bright reflexes close to the Y axis. They should be removed, then after the inverse Fourier transform the noise in the image will be much less. You can do more radical and remove all the high frequencies in the image, as shown in Fig.3. The result of the inverse transformation is shown in Fig.4.
You can change the contrast of the Fourier image: to do this, select one of the three options of the Contrast command from the Fourier menu: More (increases contrast), Less (decreases it) or Restore (restores the original setting). If you have selected some area of the spectrum, you can zero its inner or outer part. To do this, there are commands Zero Inside and Zero Outside, available by right-clicking and also from the Fourier menu. Similarly, you can increase or decrease the contribution of selected frequencies using the Increase or Decrease commands.
By amplifying, attenuating and zeroing the selected frequencies, the changes in the Fourier spectrum are immediately reflected in the surface image. Weakening of the low-frequency component in the image results in removal of macro relief.
Histogram. Histogram is often a good way to quickly estimate the height of film or other uniform objects. After executing the Histogram command, a window containing distribution of surface image points by heights appears. When moving dashed vertical lines in the bottom line of the window, numerical data are displayed (Fig.5):
Х0 – is the surface height corresponding to the left mark on the histogram,
Х1 – is the surface height corresponding to the right mark on the histogram,
Х1–Х0 – is the difference in heights between these two levels,
% – percentage of the number of surface points whose height is between the selected heights X0 and X1.
Fig.5 shows the histogram of the image X of potato virus: there is a clear left peak corresponding to the distribution of pure substrate points, the right peak (small) characterises the distribution of particles by height.
Using the Crop command from the Histogram menu, the height of each point outside the vertical lines becomes equal to the height indicated by the corresponding vertical line. These changes are immediately reflected in the surface image. The histogram is also rebuilt again and can have a completely new appearance.
CONCLUSIONS
Atomic force microscopy (AFM) and high-speed scanning have greatly advanced the study of biological objects and structures from single molecules to the cellular level. To facilitate interpretation of limited resolution images, data processing and filtering play an increasingly important role in understanding AFM measurements. When software can be used to characterise the surface topography features of a sample, its structure, and calculate geometric features, it greatly complements the computational analysis of objects.
FemtoScan Online software, along with a user-friendly interface for image processing and filtering, allows to obtain many quantitative features thanks to clever algorithms and high-quality visualisation, which certainly contributes to the understanding of processes beyond topographic images [3]. This work illustrates the capabilities of FemtoScan Online software and emphasises the importance of AFM data processing to complement experimental observations.
ACKNOWLEDGMENTS
The work of the staff of the Physical Department was carried out under the state order with the financial support of the Physical Department of the Lomonosov Moscow State University (Registration subject 122091200048-7). FemtoScan Online software is provided by Advanced Technologies Center (www.nanoscopy.ru).
PEER REVIEW INFO
Editorial board thanks the anonymous reviewer(s) for their contribution to the peer review of this work. It is also grateful for their consent to publish papers on the journal’s website and SEL eLibrary eLIBRARY.RU.
Declaration of Competing Interest. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Readers feedback
