Benjamin_Loison/Robust_image_source_identification_on_modern_smartphones
1
0
Fork 0
Code Issues 68 Pull Requests Packages Projects Releases Wiki Activity

Use mean as denoiser #57

New Issue
Open
opened 2024-04-23 04:50:30 +02:00 by Benjamin_Loison · 2 comments
Benjamin_Loison commented 2024-04-23 04:50:30 +02:00
Owner
Copy Link

RAISE flat field:

multiple_colors

Modified brightness and contrast:

image

Being able to explain why this method leads to an estimated PRNU with more color gradients as the original image.

RAISE flat field: ![multiple_colors](/attachments/f1f1fabb-971b-4e0d-afc8-af0bbebe1d01) Modified brightness and contrast: ![image](/attachments/25c31e4b-e3ae-4194-96db-7700bee58779) Being able to explain why this method leads to an estimated PRNU with more color gradients as the original image.
image.png
936 KiB
multiple_colors.png
9.8 MiB
Benjamin_Loison added the
enhancement
medium priority
medium
 labels 2024-04-23 04:50:30 +02:00
Benjamin_Loison pinned this 2024-04-23 04:50:49 +02:00
Benjamin_Loison commented 2024-04-24 14:20:08 +02:00
Author
Owner
Copy Link

Current issue is that in theory extracting noise by substracting the average image removes the PRNU, but as shown above we still observe patterns looking like unique artifacts. To avoid removing the PRNU as the image are quite smooth enough we will apply a Gaussian kernel to blur the image to remove the PRNU from the average image.

The good question is what sigma value is appropriate? Once will have mathematical metrics could benchmark.

Current issue is that in theory extracting noise by substracting the average image removes the PRNU, but as shown above we still observe patterns looking like unique artifacts. To avoid removing the PRNU as the image are quite smooth enough we will apply a Gaussian kernel to blur the image to remove the PRNU from the average image. The good question is what `sigma` value is appropriate? Once will have mathematical metrics could benchmark.
Benjamin_Loison commented 2024-04-27 14:57:43 +02:00
Author
Owner
Copy Link

Does not seem as precise as bilateral denoiser.

Mean denoiser:

image

Bilateral denoiser:

image

Source: issues/49#issuecomment-1693

Does not seem as precise as bilateral denoiser. Mean denoiser: ![image](/attachments/da53cc48-467e-4a33-bab7-803ad3e35ce9) Bilateral denoiser: ![image](/attachments/a9f58d89-613f-4f5e-9cab-cb5b6ee0ba7f) Source: [issues/49#issuecomment-1693](https://gitea.lemnoslife.com/Benjamin_Loison/Robust_image_source_identification_on_modern_smartphones/issues/49#issuecomment-1693)
image.png
6.1 MiB
image.png
4.8 MiB
Benjamin_Loison unpinned this 2024-04-27 14:58:02 +02:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
Something is not working
Context-Adaptive Interpolator
duplicate
This issue or pull request already exists
enhancement
New feature
epic
A task that is going to take more than a day to complete.
help wanted
Need some help
high priority
invalid
Something is wrong
left for future work
low priority
medium
A task that should take less than a day to complete.
medium priority
meta
question
More information is needed
quick
A task that should take less than two hours to complete.
wontfix
This won't be fixed
No Label
enhancement
medium
medium priority
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Benjamin_Loison/Robust_image_source_identification_on_modern_smartphones#57
No description provided.
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?