Shouldn't the value received by adding all the numbers (after multiplication with image matrix and kernel matrix) be put at the center of the matrix i.e., at the position where '5' is there
keras.layers.Conv2D(32, (3, 3)) uses 32 filters, right? keras.layers.Conv2D(256, (3, 3)) even uses 256 filters that are all the same? How do they look like? are they structured like this: [[ 1 0 1 ] [ 0 1 0 ] [ 1 0 1 ]] how can i choose different filters. why use many filters in the first place? isnt one enough? this is pretty cryptic to me and i dont find any pages on the web on this topic
Hello, I have an 11×11 image having in its center a 5×5 square, the image it's noiseless and I don't know how to compute the gradient of the image function given by the compass operator for this image. If I remember correctly, I should use a derivative, but I don't know exactly what and how can I use it. Can you please help me?
I am using X as a filter that is 1 for X and 0 for other pixels. My filter is a 5 by 5 filter. My image is also an X but my output is coming only a circle at the center of the image. I am confused why? please can you help?
Respected Sir, Thanks for this video with a detailed explanation. Now, can you make a video on the Cross-Correlation formula? And also explain in it how it works?
sorry can i ask you something, related to this video, but no so related actualy ahahaha I use Quantum GIS for process my image, i want to use majority filtering, but there is a problem, they want me to choose a square or a circle for the choose mode (one of the paramter, you can check all the parameter here: www.saga-gis.org/saga_tool_doc/2.2.0/grid_filter_6.html) what i want to ask is, what is search mode? how circle work and square either I think this tool use convolution thanks
I regret to tell you this that I find myself totally flung on this one. I'm really not sure. I tought this topic at undergraduate level once and never got a chance to bump into it again :)
![[Free] Guitar Fretboard Chart with notes Frequencies and Octaves](http://i.ytimg.com/vi/jL2k2pCXVAA/mqdefault.jpg)
![[Free] Guitar Fretboard Chart with notes Frequencies and Octaves](/img/tr.png)
Awesome! Great explanation, thanks!
Thanks for the easy explanation!
very good explanation. Thanks a lot.
Thank you..❤
Shouldn't the value received by adding all the numbers (after multiplication with image matrix and kernel matrix) be put at the center of the matrix i.e., at the position where '5' is there
Nope mam , it's placed on the left hand top corner . See my subsequent video for visual presentation of this.
Yes, even I've the same question. Because output of convolution is same as image size. Also, by this way you won't be able to convolve last 2 columns!
that is right, this video is wrong..
@@kshitijchhajed3242 Output size is not automatically the same as input. You can make it so by padding, however the video is indeed correct.
Thank you Nipun!
keras.layers.Conv2D(32, (3, 3)) uses 32 filters, right?
keras.layers.Conv2D(256, (3, 3)) even uses 256 filters that are all the same? How do they look like? are they structured like this: [[ 1 0 1 ] [ 0 1 0 ] [ 1 0 1 ]] how can i choose different filters.
why use many filters in the first place? isnt one enough?
this is pretty cryptic to me and i dont find any pages on the web on this topic
Informative Video!
Hello, I have an 11×11 image having in its center a 5×5 square, the image it's noiseless and I don't know how to compute the gradient of the image function given by the compass operator for this image. If I remember correctly, I should use a derivative, but I don't know exactly what and how can I use it. Can you please help me?
what would happen if used a 1x3 kernel instead of a 3x3
thanks sir, you help me for the morning test
Glad to hear that
Great explanation
Glad you liked it
Great 👏
Appreciate that !!
I am using X as a filter that is 1 for X and 0 for other pixels. My filter is a 5 by 5 filter. My image is also an X but my output is coming only a circle at the center of the image. I am confused why? please can you help?
I'm sorry I didn't comprehend
How can I see the 19 lectures before
This is informative video
thankyou so much
What is the difference between kernal and filter size?
Kernel is 2D arrangement of weights and when a few Kernels together perform convolution or in 3D ,its a filter. So a 2D filter and a kernel is same
Respected Sir,
Thanks for this video with a detailed explanation.
Now, can you make a video on the Cross-Correlation formula? And also explain in it how it works?
Thanks and I will surely try to make one video on that soon :)
@@nipunharitash Ok Sir. I am waiting for that.
sorry can i ask you something, related to this video, but no so related actualy ahahaha
I use Quantum GIS for process my image, i want to use majority filtering, but there is a problem, they want me to choose a square or a circle for the choose mode (one of the paramter, you can check all the parameter here: www.saga-gis.org/saga_tool_doc/2.2.0/grid_filter_6.html)
what i want to ask is, what is search mode? how circle work and square either
I think this tool use convolution
thanks
I regret to tell you this that I find myself totally flung on this one. I'm really not sure. I tought this topic at undergraduate level once and never got a chance to bump into it again :)
Oh okey noprob, btw your video teach me well :D keep it up
how to code this filter in python man?
will try