| Summary: | Duplicate Detection Reference Image | ||
|---|---|---|---|
| Product: | [Applications] digikam | Reporter: | Ron <ronaday> |
| Component: | Searches-Similarity | Assignee: | Digikam Developers <digikam-bugs-null> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | caulier.gilles, martin.marmsoler, metzpinguin |
| Priority: | NOR | ||
| Version First Reported In: | 8.0.0 | ||
| Target Milestone: | --- | ||
| Platform: | Microsoft Windows | ||
| OS: | Microsoft Windows | ||
| Latest Commit: | https://invent.kde.org/graphics/digikam/commit/fecc957fda73704b05c425dbacaa47a42319228c | Version Fixed/Implemented In: | 8.2.0 |
| Sentry Crash Report: | |||
|
Description
Ron
2023-03-06 12:11:07 UTC
Is it really digiKam-8.0.0, is it correct? It is valid for older versions, but not for 8.0.0. Not only the date is used to determine the original image, but also the file size and image size. I've tested it with many sample images and the "better" image is always set as the reference image. Maik Git commit fecc957fda73704b05c425dbacaa47a42319228c by Maik Qualmann. Committed on 06/03/2023 at 20:18. Pushed by mqualmann into branch 'master'. more priority on pixel and file size for reference image FIXED-IN: 8.0.0 M +1 -1 NEWS M +7 -7 core/libs/database/haar/haariface.cpp https://invent.kde.org/graphics/digikam/commit/fecc957fda73704b05c425dbacaa47a42319228c I am running version 8. This is ->not<- fixed! Why are you lieing? Why not just admit it's not fixed? The Windows installer 8.0.0 pre-release do not integrate yet the last commit in source code from comment #2 by Maik. This will fix your problem. You needs to wait a more recent rebuild of 8.0.0 windows installer available here : https://files.kde.org/digikam/unstable/ Gilles Caulier Hi Ron, 8.0.0 is published. This report is fixed now ? Best regards Gilles Caulier I can have a look into this, because I worked a few days in this area so I am familiar with the code Ron, can you give me sample images where this happens? You can also try to download some from the internet and crop them to get a simple reproducable example. So it is easier to fix the issue I'm pretty sure the bug is fixed. I will test it again tonight with the new implementation. Ron tested a version in Comment 3 that did not yet contain the fix from Comment 2. Maik Well... fixed or not fixed, that the Q (:=-))))... Gilles |