[feature] sift: update default values, add new filtering and add dsp-sift variation #935
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
firstScale is changed only based on input image resolution
Use a new explicit parameter instead of to allow multi-threading on descriptor extraction.
- add NonExtremaFiltering - numOctaves: use auto instead of fixed value - firstOctave is change when the image is small
And should thus be called explicitely before using other vlfeat functions.
fcastan
force-pushed
the
dev/siftGridSort
branch
from
4f6a1b2
to
343c64c
Compare
ghost
reviewed
Nov 26, 2020
ghost
previously approved these changes
Nov 27, 2020
| frameAtScale.a21 *= dspScale; | ||
| frameAtScale.a22 *= dspScale; | ||
|
|
||
| vl_covdet_extract_patch_for_frame(covdet.get(), patch.data(), kPatchResolution, &internalBuffer, |
There was a problem hiding this comment.
To me, this is extracting a sub image at a given location at an interpolated scale.
I assume the scale interpolation is based on the existing, pre-computed scales. So this is bounded by the parameters of the sift detector. As such, shouldn't the scales be computed to make sure that for each keypoint, we are able to get a range of scales large enough ?
simogasp
requested changes
Nov 27, 2020
Comment on lines
+28
to
+29
| case feature::EImageDescriberType::DSPSIFT: return "yellow"; | ||
|
|
There was a problem hiding this comment.
Suggested change
| case feature::EImageDescriberType::DSPSIFT: return "yellow"; | |
| case feature::EImageDescriberType::DSPSIFT: return "yellow"; |
There was a problem hiding this comment.
I do not see the change.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Feature point quality has a critical impact on SfM and Panorama pipelines.
This PR re-evaluate the default parameters and add new filtering strategies to be less impacted by the peak threshold in standard sift. This parameter is critical in particular on HDR images.
Features list
Implementation remarks