Professional Documents
Culture Documents
Group members:
1. Estifanos Alebachew …….SGS/0231/2014A
2. Mulatu Tadele ……………SGS/0237/2014A
3. Tatek Shenkute …………..SGS/0242/2014A
4. Tilahun Brehanu …………SGS/0393/2013A
Abstract
Images are often degraded by noises. Noise can occur and obtained during image capture, transmission and
so on. Noise removal is an important task in image processing. In general the results of the noise removal
has a strong influence on the quality of the image processing techniques. Several techniques for noise
removal are well established in color image processing. The nature of the noise removal problem depends
on the type of the noise corrupting the image. In this Project we are going to See Two kind of noise Poisson
noise and spackle noise.
And also we are trying to apply different methods such as Gaussian filtering, mean filtering and median
filtering algorithms for reducing the noise. Also we are using the ranking algorithms which are PSNR and
MSN to see the better results for filtering technique. For as according to the PNSR values of the spackle
noise median filtering is better than other filtering technique the same thing in Poisson filtering we get a
best result in mean filtering technique.
1|Page
2|Page
One of the fundamental challenges in image processing and computer vision is image de-noising.
What de-noising does is to estimate the original image by suppressing noise from the image. While
in digital image processing there always comes an integrated part which is Noise. Noise means
random disturbance in a signal in a computer version. In our case, the signal is an image. Random
disturbance in the brightness and color of an image is called Image noise. Image noise is a random
variation of brightness or color information in the images captured. This is the deterioration of the
image signal due to an external source. Images with multiplicative noise are characterized by the
fact that brighter areas contain more noise.
In this project, we consider two kinds of noise speckle noise and Poisson noise. As we learn in the
class Speckle Noise is a granular 'noise' that inherently exists in and degrades the quality of the
image. Speckle noise also occurs caused by the coherent processing of backscattered signals from
multiple distributed targets. And in other words, speckle noise can be modeled by random values
multiplied by pixel values of an image. And in medical image processing, Image de-noising plays a
very important role in diagnosis. Medical images are very much inconsistent. In medical imaging,
Ultrasound is very popular due to its low cost, least harmful to the human body, real-time system,
and being small in size. Besides its advantages, it suffers from a major problem, speckle noise in
ultrasound images Thus, this project attempt to identify the effect of Poisson and Speckle noise on
the original image and try to use the Given two filtering algorithm Median and mean and one of our
selecting filtering techniques for each noise. In addition, by using MSE and PSNR comparison of
each filtering technique result is presented.
2|Page
3|Page
1. 2. Speckle noise
Speckle noise: removing of noise from images is now a very challenging issue in the field of image processing,
this paper attempts to undertake the study of speckle noise of image by using three types of filters as mean filter,
median filter applied in the four types of image and compared with one another so as to choose the best technique.
In the project we used different types of algorithms in order to de-Nosie Speckle and Poisson from the
image.
Median filtering, Mean and Gaussian Filtering.
Advantages:
1 .Median filter is more robust than mean filter so a single very unrepresentative pixel in a neighborhood
will not affect the median value significantly
3|Page
4|Page
The blur function from the open-CV library can be used to apply a mean filter to an image. New_image
=cv2.blur(image,(firure_size,figure_size))
4|Page
5|Page
3. Experimental results
3.1 adding a Poisson noise to the image, filtering and its result
5|Page
6|Page
6|Page
7|Page
3.1 adding a spackle noise to the image, filtering and its result
Img 2.1. Original Image Img 2.2. Image With spackle noise
7|Page
8|Page
8|Page
9|Page
3.3. Experimental Image result values for spackle noise with a colored image
Filtering technique used MSE and PNSR values
Mean filtering is an effective noise reduction technique. As we can see in the picture 1.4, the picture is
better than others. We assume that this filtering technique is better than others for Poisson filtering.
Gaussian Image Filtering 1.3 is the second effective filtering for Poisson noise filtering. But the image is a
bit blurry. So we assume it works fine for Poisson noise but not better than mean filtering. In Img: 1.4
median filtering, it is not de-noised the noise, so we conclude that this filtering technique is not working
in poisoned noise.
In this noise we are trying to apply three filtering techniques which is mean, median and Gaussian filtering.
Regarding the result of the experiment that is by using MSE for comparison, from the 3 speckle noise filters
shown in the above pictures and the table below, we find that the median filter is best suited filter than the
rest of the other filters to reduce speckle noise.
9|Page
10 | P a g e
5. Conclusion
Noise is always present in digital images during the image acquisition, encoding, transmission, and
processing stages. In this paper, Poisson noise and speckle noise are used, and various filtering
algorithms are applied to reduce noise.
Image filters can be used to reduce the amount of noise in an image and enhance the edges of an
image. As we have seen in the experimental results Mean filtering are best for Poisson noise and
median and mean filtering are best for speckle noise.
10 | P a g e
11 | P a g e
Reference:
[1]. Jyoti J. and and Rajveer S. (2015). A STUDY OF SPECKLE NOISE REDUCTION FILTERS.
Signal & Image Processing: An International Journal (SIPIJ) Vol.6
[2]. Er. K.B. and Er. A. (2014). A Review on Speckle Noise Reduction Techniques. IOSR
Journal of Computer Engineering (IOSR-JCE), 2278- 8727Volume 16, Issue 3, Ver. I.
[4]. https://github.com/bnsreenu/python_for_image_processing_APEER
11 | P a g e
12 | P a g e
Members Contribution
Tatek Shenkute and In spackle noise adding and applying different filtering
Tilahun Brehanu techniques.
12 | P a g e