Hi Edgar, I'm sorry for bothering you again. I'm doing some study on coordinate conversion and find out that there is another way to decide the resampling location which is calculate the center location first. once we get the center location , then we can calculate the pixel distance to center location and multiply the distance by scaling factor to form a new position. for example , if center location is x=4 , then X' = (x-4) * scaling factor . do you have any opinion for this kind of resampling method? does it matched the alignment you talk in the video ? nice to see your feedback.
Hi, Thanks for your sharing . It's very helpful :) However I'm so confusing when you are talking about " to align with the default point" begins in the video 2:30 . I'm just not sure what exactly "Align" means and how does it work. Here is my thought ( should be wrong) for interpolating from 3 point into 4 points. Initially we have X= 0 , 1 , 2 (3 points, 2 intervals) , then why don't we just fixed the boundary X=0 and X=2, then split the whole intervals into 3 equally distance parts ? ( 4 points, 3 intervals) If we do that, then we should have X=0 , X=2/3 , X=4/3, X=2 for new 4 points.(instead of X= -0.125, 0.625,1.375,2.125) and The Y(gray value) will be 155, 173, 176,163 respectively. Could you shed some lights on it ? Thanks for your kindly response.
Yes, you are right. It is another form of linear interpolation. But the interpolated images won't look aligned (compared to the original image) if you do it that way. You can do the test. Thank you for your question.
Hi Edgar, I'm sorry for bothering you again. I'm doing some study on coordinate conversion and find out that there is another way to decide the resampling location which is calculate the center location first. once we get the center location , then we can calculate the pixel distance to center location and multiply the distance by scaling factor to form a new position. for example , if center location is x=4 , then X' = (x-4) * scaling factor . do you have any opinion for this kind of resampling method? does it matched the alignment you talk in the video ? nice to see your feedback.
Lovely video thank you very much!
🙏
Hi, Thanks for your sharing . It's very helpful :)
However I'm so confusing when you are talking about " to align with the default point" begins in the video 2:30 .
I'm just not sure what exactly "Align" means and how does it work.
Here is my thought ( should be wrong) for interpolating from 3 point into 4 points.
Initially we have X= 0 , 1 , 2 (3 points, 2 intervals) , then why don't we just fixed the boundary X=0 and X=2, then split the whole intervals into 3 equally distance parts ? ( 4 points, 3 intervals)
If we do that, then we should have X=0 , X=2/3 , X=4/3, X=2 for new 4 points.(instead of X= -0.125, 0.625,1.375,2.125) and The Y(gray value) will be 155, 173, 176,163 respectively.
Could you shed some lights on it ?
Thanks for your kindly response.
Yes, you are right. It is another form of linear interpolation. But the interpolated images won't look aligned (compared to the original image) if you do it that way. You can do the test.
Thank you for your question.
very helpful :))
Great thank you for your video 😊
👍 Thank you
You're Chopping it
How we can find 0,625 | 1,375 | 2,125 ? I don't understand this, please help me about that
You can just simple solve the equation by assign n from 0 to 3
gave you a thumbs up, but next time please narrate the video by a human and not a machine. Thanks.
🙏
P-1, P-2. ULTRAKILL brain rott