Image organizer

Not so common, or differentiating features

• Pictures can be organized by one or more mechanisms
  • Images can be organized into folders, which may correspond to file-system folders.
  • Images may be organized into albums, which may be distinct from folders or file-system folders.
  • Albums may be organized into collections, which may not be the same as a folder hierarchy.
  • Grouping or sorting by date, location, and special photographic metadata such as exposure or f-stops if that information is available. See Exif for example.
  • Images can appear in more than one album
  • Albums can appear in more than one collection
  • Grouped or stacking of images within an album, by date, time, and linking copies to originals.
  • Adding and editing titles and captions
• Simple or sophisticated search engines to find photos
  • Searching by keywords, caption text, metadata, dates, location or title
  • Searching with logical operators and fields, such as "(Title contains birthday) and (keywords contain cake) not (date before 2007)"
• Separate backing up and exporting of metadata associated with photos.
• Retouching of images (either destructively or non-destructively)
• Editing images in third-party graphical software and then re-incorporating them into the album automatically
• Stitching to knit together panoramic or tiled photos
• Grouping of images to form a slideshow view
• Exporting of slideshows as HTML or Flash presentations for web deployment
• Synchronizing of albums with web-based counterparts, either third-party (such as Flickr), or application specific (such as Lightroom or Phase One Media Pro).
• Retention of Exif, IPTC and XMP metadata already embedded in the image file itself

Two categories of image organizers

• Automatic image organizers. These are software packages that read data present in digital pictures and use this data to automatically create an organization structure. Each digital picture contains information about the date when the picture was taken. It is this piece of information that serves as the basis for automatic picture organization. The user usually has little or no control over the automatically created organization structure. Some tools create this structure on the hard drive (physical structure), while other tools create a virtual structure (it exists only within the tool).
• Manual image organizers. This kind of software provides a direct view of the folders present on a user's hard disk. Sometimes referred to as image viewers, they allow the user only to see the pictures but do not provide any automatic organization features. They give maximum flexibility to a user and show exactly what the user has created on their hard drive. While they provide maximum flexibility, manual organizers rely on the user to have their own method to organize their pictures. Currently there are two main methods for organizing pictures manually: tag and folder based methods. While not mutually exclusive, these methods are different in purposes, procedures, and outcomes.

Many commercial image organizers offer both automatic and manual image organization features. A comparison of image viewers reveals that many free software packages are available that offer most of the organization features available in commercial software.

Future of image organization

There are several imminent advances anticipated in the image organization domain which may soon allow widespread automatic assignment of keywords or image clustering based on image content:

• colour, shape and texture recognition (For example, Picasa experimentally allows searching for photos with primary colour names.)
• subject recognition
• fully or semi-automated facial, torso or body recognition (For example, FXPAL in Palo Alto experimentally extracts faces from images and measures the distance between each face and a template.)
• geo-temporal sorting and event clustering. Many software will sort by time or place; experimental software has been used to predict special events such as birthdays based on geo-temporal clustering.

In general, these methods either:

• automatically assign keywords based on content, or
• measure the distance between an untagged image and some template image which is associated with a keyword, and then propose that the operator apply the same keyword(s) to the untagged images

Notable image organizers