I'v seen difference between the equation applied in the script and the gaussian equation. the kernel (i,j) in script should be divided also by 2*pi*sigma^2 is it right???
It's not necessary. Since that is a constant and it will not matter at the end, because the intensities are normalized at the end (to bring them on 0~255 scale)
Thanks for clearing, and for the code. I have a question, what is the difference between applying your code and the Gaussian code implemented in Matlab (g1=mat2gray(fspecial('gaussian',256,1))?
Could you please help me get some sort of grasp on the following question, it would be of great help mate... Implement your own Matlab function that performs a 9x9 Gaussian filtering. Your function interface is: my_Gauss_filter( noisy_image, my_9x9_gausskernel, output_image) NOTE: In doing this task you MUST NOT use any Matlab's in-built image filtering functions (e.g. imfilter(), filter2(), conv2(), or filter(),conv()). Inotherwords,youarerequiredtocode your own 2D filtering matlab code, based on the original mathematical definition for 2D convolution. However, you are allowed to use Matlab's function fspecial() to generate a 9x9 sized Gaussian kernel. Apply your Gaussian filter to the above two noisy images, and display the resultedimagesand visually check their noise-removal effects. (Note: One of key parameters to choose for the task of image filtering is the standard deviation of your Gaussian filter. You may need to test and compare different Gaussian kernels with different standard deviations.)
Hi Alan, sorry for late reply. I think my code does correctly implement your question's requirement. I tried my best to summarize the working of this filter in my video but could not cover the very basic in such short video. First you can access my code here : pastebin.com/SKNGrFwb But here's what will be different in your code than what I've done in video : (i) You have to implement 9x9 filter. What I did was 5x5 filter. So change *** Im = padarray(I,[2 2]); *** to *** Im = padarray(I,[4 4]); *** change *** temp = Im(i:i+4 , j:j+4); *** to *** temp = Im(i:i+8 , j:j+8); *** (ii) You can use produce gaussian kernel of 9x9 size using builtin function. (refer to matlab documentation). So you can basically leave the part between these 2 comments *** % making a gaussian kernel *** and ***% now we apply the filter to the image *** (iii) I just glanced over the code so you might need to look for some changes that might need to be done, so just check if output is desired or not. If you need to understand the basics of how this filter works. then check out my reply on comment by @akshay ts. I am posting link of the book : mega.nz/#!kjxTXKwJ!_vCH770JV_apmwNvmclSFAXYvH4UX0qpX-MNi-O9GDQ Just read sections 3.4 , 3.5 and 3.6 and that will cover your theory part for mean filter and then watch few starting minutes of this video again and you will get it. It is just 20 or so pages of trivial reading, nothing fancy. If you have any questions after this, please reply or drop an email at geekbitofeverything@gmail.com Hope this helps.
Hey! Thanks for the reply dude! Don't you mean: temp = Im(i:i+8 , j:j+8);? I took my time to understand what was actually happening inside the iteration blocks and finally got it to work, albeit it was a few hours. But hey, I literally just started image processing :P Couldn't have done it without you mate! :D
Yes, sorry for the typing error. I copied last text to change it but forgot. Glad you caught it. And glad to help man. I will try some of next videos to try and cover more basic ground and adding audio too, which will be much more helpful. Just need to get me some free time. If you need any help, even on some other Image Processing concepts off this video, drop me a mail at geekbitofeverything@gmail.com I will reply unless I am swamped with such email which is improbable in near future :)
Sorry about that man. I thought no audio was better than what I recorded on my system. I don't know what you have covered in Image processing so far, so I am uploading a book at this link : mega.nz/#!kjxTXKwJ!_vCH770JV_apmwNvmclSFAXYvH4UX0qpX-MNi-O9GDQ Just read sections 3.4 , 3.5 and 3.6 and that will cover your theory part. If you understand the mean filter (I am assuming you don;t since you are new to this) then you will get Gaussian filter easily. Don't read all the chapter if you know what pixels and pixel densities are. Ask any doubt you have, I will try to reply as soon as possible.
Sir y u have taken that square distance as (i-3) like that
Great! Many thanks!!
I'v seen difference between the equation applied in the script and the gaussian equation.
the kernel (i,j) in script should be divided also by 2*pi*sigma^2 is it right???
It's not necessary. Since that is a constant and it will not matter at the end, because the intensities are normalized at the end (to bring them on 0~255 scale)
Thanks for clearing, and for the code.
I have a question, what is the difference between applying your code and the Gaussian code implemented in Matlab (g1=mat2gray(fspecial('gaussian',256,1))?
@@qutaibahallak8354 Nothing. It's about learning and also most professors don't allow that in college.
for 3*3 kernel, is the same?
Is this a low pass or a high pass filter?
lowpass.... blurring filters are lowpass and sharpening filters are highpass filters
can I get the for gaussian blurring ?
hye geek. do u know code for distance detection ?
please can you write code gaussian filter in information box or in comment because the vedio not clare and not sound
Could you please help me get some sort of grasp on the following question, it would be of great help mate...
Implement your own Matlab function that performs a 9x9 Gaussian filtering. Your function interface is: my_Gauss_filter( noisy_image, my_9x9_gausskernel, output_image) NOTE: In doing this task you MUST NOT use any Matlab's in-built image filtering functions (e.g. imfilter(), filter2(), conv2(), or filter(),conv()). Inotherwords,youarerequiredtocode your own 2D filtering matlab code, based on the original mathematical definition for 2D convolution. However, you are allowed to use Matlab's function fspecial() to generate a 9x9 sized Gaussian kernel. Apply your Gaussian filter to the above two noisy images, and display the resultedimagesand visually check their noise-removal effects. (Note: One of key parameters to choose for the task of image filtering is the standard deviation of your Gaussian filter. You may need to test and compare different Gaussian kernels with different standard deviations.)
Hi Alan, sorry for late reply. I think my code does correctly implement your question's requirement. I tried my best to summarize the working of this filter in my video but could not cover the very basic in such short video.
First you can access my code here : pastebin.com/SKNGrFwb
But here's what will be different in your code than what I've done in video :
(i) You have to implement 9x9 filter. What I did was 5x5 filter. So
change *** Im = padarray(I,[2 2]); *** to *** Im = padarray(I,[4 4]); ***
change *** temp = Im(i:i+4 , j:j+4); *** to *** temp = Im(i:i+8 , j:j+8); ***
(ii) You can use produce gaussian kernel of 9x9 size using builtin function. (refer to matlab documentation). So you can basically leave the part between these 2 comments *** % making a gaussian kernel *** and ***% now we apply the filter to the image ***
(iii) I just glanced over the code so you might need to look for some changes that might need to be done, so just check if output is desired or not.
If you need to understand the basics of how this filter works. then check out my reply on comment by @akshay ts.
I am posting link of the book :
mega.nz/#!kjxTXKwJ!_vCH770JV_apmwNvmclSFAXYvH4UX0qpX-MNi-O9GDQ
Just read sections 3.4 , 3.5 and 3.6 and that will cover your theory part for mean filter and then watch few starting minutes of this video again and you will get it. It is just 20 or so pages of trivial reading, nothing fancy.
If you have any questions after this, please reply or drop an email at geekbitofeverything@gmail.com
Hope this helps.
Hey! Thanks for the reply dude!
Don't you mean: temp = Im(i:i+8 , j:j+8);?
I took my time to understand what was actually happening inside the iteration blocks and finally got it to work, albeit it was a few hours. But hey, I literally just started image processing :P
Couldn't have done it without you mate! :D
Yes, sorry for the typing error. I copied last text to change it but forgot. Glad you caught it.
And glad to help man. I will try some of next videos to try and cover more basic ground and adding audio too, which will be much more helpful. Just need to get me some free time.
If you need any help, even on some other Image Processing concepts off this video, drop me a mail at geekbitofeverything@gmail.com
I will reply unless I am swamped with such email which is improbable in near future :)
How to make a 3x3 gaussian filter? what changes I need to make in this code?
sir for 9*9 gausian filter where will the change occur?
a trick: you can watch movies at flixzone. Me and my gf have been using it for watching lots of of movies lately.
@Lorenzo Shawn Definitely, have been watching on Flixzone for months myself =)
hey guys, how can we download the code?
pastebin.com/SKNGrFwb
please post the code . i am trying now
link is mentioned in replies of some comments above
Can I get an explanation?
Do you mean in general or you have a specific question? Just ask away man.
Geek Bit of Everything I am a novice... It would be helpful if it has audio embedded with this video.. just needed to do my own Gaussian filter...
Sorry about that man. I thought no audio was better than what I recorded on my system.
I don't know what you have covered in Image processing so far, so I am uploading a book at this link :
mega.nz/#!kjxTXKwJ!_vCH770JV_apmwNvmclSFAXYvH4UX0qpX-MNi-O9GDQ
Just read sections 3.4 , 3.5 and 3.6 and that will cover your theory part. If you understand the mean filter (I am assuming you don;t since you are new to this) then you will get Gaussian filter easily.
Don't read all the chapter if you know what pixels and pixel densities are. Ask any doubt you have, I will try to reply as soon as possible.
Thanks.... Can I have your personal email?
geekbitofeverything@gmail.com
My all email accounts are linked to single one, so I check this everyday.