![]() ![]() There probably won't be tags at first (unless you've created or imported some), but you can create tags in the Tag Manager interface and assign them to pictures. Tags filters the photos in your collection by arbitrary metadata tags.In addition to the Album view, there are: There are lots of other ways to view your photos, though, and they're all accessible as verticle tabs (a time-honored KDE interface tradition) along the left edge of the main window. When you select a directory (or an album, if you prefer), the photos within are displayed as an array in the right panel. DigiKam refers to whatever directories it finds within your image directory as an Album, and it parses each image file, along with its native metadata plus metadata digiKam allows you to add, into a thumbnail view. The initial default view is a filesytem view, starting from whatever directory you defined as your image folder during digiKam setup. ![]() The basics are pretty simple: On the left are panels that control how you view photos, in the middle are the photos themselves, and on the left are effects and filters. The layout of digiKam is fairly intuitive, especially if you're a KDE user. Depending on how many photos you have and how large they are, you might want to let it run over night. The initial launch will be slower than usual because digiKam must analyse the photographs in your collection and record information about each one. Some choices affect performance and file size, so read the screen carefully to decide what you really want, but all of the decisions can be changed later, so accepting the defaults is safe if you're not sure. Upon first launch, you must step through a brief setup wizard. Your distribution may or may not have the latest version, but don't get overly concerned about that digiKam is in the enviable position of having been essentially a complete and stable application for years now, so unless you're looking for a specific feature that only exists in the latest version, it's going to be a good experience, at least just getting started. If you don't have digiKam installed, you can either grab it for Linux, Windows, or, with a little bit of work, on OS X from the digiKam download page or, on Linux, from your distribution's software repository. Suffice it to say that digiKam is a humble application, because "digital photo management" barely touches on the feature set. Welcome to the communityĭigiKam bills itself as a digital photo management application.I guess is good to do it from time to time when new people is added to your collection and/or when you have matched a lot of faces and people, specially if you have fixed many mismatched faces. What I don't know is when you need to re-train the database. IMPORTANT: On the settings sugbtab, select the YOLO v3 detection model. ![]() I guess this 'trains' the database using the faces already detected and assigned.įinally, when you hace the database trained, you can go to the People tab again, select "Recognize faces" and let Digikam match detected faces with people. Go to Tools > Maintenance and select the update faces. But first, and this seems to be key, you have to instruct the app the train the faces recognition database. ![]() Once you have a minimum set of faces, you can use the recognize feature. Once you have faces detected assign each faces with a person. I would start with a limited set of albums. On the workflow subtab (at the bottom), select "Detect faces". How to detect and assign faces? You can do it manually (editing a picture and adding a face), or you can go to the People tab and use the detect faces feature. If not, you will need to do some more work. So, you need some faces already in place (this means detected and assigned to someone). I'm not totally sure how it works, but AFAK the feature needs some training. On first start up, application needs to download some external data for the face recognition. First one is to detect faces(detect parts of the picture that are faces), and the second one is to recognize the faces (assign a detected face to someone). Keep in mind that this woprks in two phases. It seems Digikam 7.2 improved this feature. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |