GRK

Automation of DIP-Coating ZnO thin films and its structural studies
Paper presented at International Conference on Nano Electronic science

& Technology
Sri Vasavi College, Erode
By
G.R.Kandhasamy
Lecturer/ Selection grade
EITPC, Kavindapadi
ABSTRACT
Surface study of thin film is characterized by employing scanning
Electron microscopy (SEM). The implementation of imge processing
algorithm is to determine the pores in the SEM images of thin film.
Image segmentation is applied to detect the pores in the thin film. In this
paper, ZnO thin film is used for the surface study by various edge
detection techniques. As the reqult canny edge detection gives the best
porosity information of ZnO thin film SEM images. This result is taken
by performing the statistical analysis (ie) finding the PSNT (Peak Signal
to Noise Ratio) values. This may help to identify the particular place of
pores and pits in thin film SEM images.
Introduction
• Scanning Electron Microscope (SEM) is very useful technique to acquire the
high magnification images of the specimens. Secondary electron is predominately
used for surface property studies of the material under SEM technology. Thin film
technology is availed by many of the modern fields like MEMS, Nano technology
etc. Nowadays thin films are used to develop the solar cells. The quality of light
reflection depends on the surface of the thin films. Optical thin films are mostly
used in automotive industry for non-contact measurement of surcaces. This can be
achieved by observing the SEM images. SEM images are used to study the strain
measurement of the thin films with the help of image processing techniques. PSNR
is applied to develop the efficiency of the image segmentation results. In this
paper, we discuss the surface property of the ZnO thin film by image segmentation
(i.e. one of the image processing techniques). The objective result of the various
edge detection images are displayed and discussed in this paper.
2. MATERIALS AND METHOD
• Image processing techniques for finding the porosity consist of
Image Acquisition, Image Enhancement, Image Segmentation and
Feature extraction. The below block diagram shows the proposed
approach. The thin films were analyzed and examined using Scanning
Electron microscopy. An SEM generates high energy electrons and
focuses them on a specimen. The thin films were subjected to
secondary electron imaging in SEM produces vivid images with good
resolution 1-5 m. The images from the SEM are in indexed mode so
that these images should be converted to gray scale mode. The process
of conversion of a gray scale image into a binary is clled thresholding.
In a binary image, object pizels are represented by 1s or the highest
intensity possible in the image.
Fig 1: Block diagram of proposed method
2.1 EDGE DETECTION
2.1.1 Sobel Operator
• The operator consists of a pair of 3 x 3 convolution kernels as
shown in Figure 2. One kernel is simply the other rotated by 90º.
Fig.2: Masks used by Sobel operator

Cont...
• These kernels are designed to respond maximally to edges running
vertically and horizontally relative to the pixel grid, one kernel for each of
the two perpendicular orientations. The kernels can be applied separately
to the input image, to produce separate measurements of the gradient
component in each orientation (call these Gx and Gy). These can then be
combined together to find the absolute magnitude of the gradient at each
point and the orientation of that gradient. The gradient magnitude is given
by :
• which is much faster to compute. The angle of orientation of the edge

(relative to the pixel grid) giving rise to the spatial gradient is given by :
2.1.2 Roberts edge detection
The Roberts Cross operator performs a simple, quick to compute, 2-D spatial gradient measurement on an
image. Pixel values at each point in the output represent the estimate absolute magnitude of the spatial gradient
of the input image at that point. The operator consists of a pair of 2 x 2 convolution kernels as shown in Figure
2. One kernel is simply the other rotated by 90º. This is very similar to the Sobel operator.
2.1.3. Prewitt edge detection

Prewitt operator edge detection masks are the one of the oldest and best understood methods of detecting
edges in images. The strength of the edge at given location is then the square rote of the sum of the squares of
two derivatives.
2.1.4 Canny edge detection
The Canny edge operator takes the grayscale image as input and calculates the gradient of intensities, a
vector with both magnitude and direction. It is applied to find both the strong and weak edges of each particle.
The normal of the edge aligns with the direction of the gradient vector, and the magnitude gives the strength of
the edge. Canny edge detection uses linear filtering with a Gaussian kernel to smooth noise and then computes
the edge strength and direction for each pixel in the smoothed image. This is done by differentiating the image
in two orthogonal directions and computing the gradient magnitude as the root sum of squares of the
derivatives. The gradient direction is computed using the arctangent of the ratio of the derivatives. Edge pixels
are identified as the pixels that survive a thinning process called non-maximal suppression. In this process, the
edge strength of each edge pixel is set to zero if its edge strength is not larger than the edge strength of the two
adjacent pixels in the gradient direction.
Cont...
The following algorithms are used in canny edge detection
• Smooth the image by convolving with a Gaussian of variance.
• Compute the gradient of the smoothed image, and compute its magnitude and
direction.
• Non-maximal suppression : Select the pixels where the gradient magnitude
has a local maximum in the direction of the gradient.
• Using two specified thresholds, T1 and T2, with T1˂ T2, mark selected pixels
with gradient magnitude larger than T2 as ‘strong’ and pixels with magnitude
between T1 and T2 as ‘weak’.
• Select all strong pixels, and all weak pixels that are connected to strong pixels
horizontally, vertically or diagonally.
2.1.4. Statistical measures
• The first measurement used in image processing is the mean square
error (MSE).
• M, N is the dimensions of the image. Lower value of the MSE denotes
the less error and PSNR becomes high because the MSE has the
inverse relation with the PSNR.
• PSNR Value is high for the good result images.

3. EXPERIMENTAL WORK
The thin film has been grown from solution by dip coating method at 400ºC
temperature. The pores of the ZnO thin films are studied using image processing.
3.1 SETTINGS
Those thin films are subjected to Secondary electron imaging in scanning electron
miscroscopy under 5keV beam energy at a working distance of 500um. The SEM is saved
as graphic file format for further process in MATLAB environment. The image acquisition
is done by using ‘imread’ MATLAB command and it is converted into intensity to gray
image. The sholding is taken by converting the gray image into binary image. 2D adaptive
filter is introduced to reduce the noise in the SEM images. And this filtered image is
further processed by introducing various edge detection techniques. After applying edge
detection to the SEM image, quantitative measurement is taken using PSNR measures.
4. RESULTS
Click to adWe apply our method to SEM images. Our intention is to find the pores of
the thin films. Figure (1a) shows the graphics format of the thin film SEM image and
figure (1b) shows the filtered image of the ZnO thin films.
Figure (2a) and (2b) shows the edge detected image and corresponding pizel region
of the thin film. After filtering the image those images are subjected to edge detected
image shows the pixel regions of the pores as ones. From this result the pores of the ZnO
thin films are found using canny Click
edgetodetection.
add text
SI.NO Edge Detectors PSNR values
1. Sobel +2.38 dB
2. Roberts +2.27 dB
3. Prewitt +2.37 dB
4. Canny +2.40 dB
Table 1 : Statistical Measures

From the above table canny edge detectors gives the better results.
5. CONCLUSION
• The microstructure plays vital role in thin films. The pores of the
ZnO thin films determine the formation and growth of the thin films.
In that case detecting the pores is very important in thin films. Canny
detection gives the best result through statistical measures to find the
regions of the pores in the ZnO thin films.
THANK YOU
HAVE A NICE DAY

GRK

Uploaded by

Copyright:

Available Formats

GRK

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GRK

Uploaded by

Copyright:

Available Formats

Automation of DIP-Coating ZnO thin films and its structural studies

Paper presented at International Conference on Nano Electronic science

Fig.2: Masks used by Sobel operator

• which is much faster to compute. The angle of orientation of the edge

2.1.3. Prewitt edge detection

• PSNR Value is high for the good result images.

Table 1 : Statistical Measures

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.