Posts Tagged ‘photo-data’

From Photos to Photo-Data

March 23, 2011

Some time in the 1970’s, dedicated word processors were introduced to the world providing writers with a set of tools that enabled them to control the presentation of their words, not just the content. For the first time since Mr. Gutenberg made the pen obsolete as a publishing mechanism, writers regained a measure of control of the presentation of their words from the typesetter. Thus began a transformation where communication with written words moved from a sterile process of repeating the authors characters to a nuanced presentation that conveyed, context, emphasis and intent.

In 1983, Microsoft introduced Word. Word not only continued the evolution of the presentation tools of its earlier counterparts, but it eventually established .doc as a generally accepted standard for digital document files. And with standardization, comes efficiency, predictability and eventually innovation. Since the files in an archive of .doc’s all behave in the same way, it is possible to explore millions of documents at a time and zero-in on the nuance of a four-word phrase within seconds.

Such innovation in searching eventually led to the use of tags (or metadata) to identify details within a document, allowing for categorized searching. A practice commonly used in blogs and social media (aka Web 2.0 applications.)

Begin Search: "Ark of the Covenant"

A similar transformation is underway with digital photos. As a whole, the digital photo universe is just beginning to move beyond the storage and display of pictures on a screen to the nuanced presentation of photos complete with context, emphasis and intent. The Photo-Data era has begun.

Without tag data, photos are opaque collections of 1’s and 0’s that yield no information about the content represented by the visual images they contain. Tags offer us the ability to search through gobs of photo-data without the need for us to visually interpret each individual image.

The one drawback to tags is the tagging process itself.  Current methods for applying tags to digital photos in the consumer and professional photography markets rely on humans to review images and then select the appropriate tag(s) that describe the image content. The two problems with this process are that it is labor intensive and tag selection is subjective.  Given that there are an infinite number of tags available to describe general interest photos, and that different photos mean different things to different people, it would be next to impossible to develop an automated method for effectively tagging all photos.

However, if we instead consider a group of photos from a defined audience with a shared perspective (with variances), the tagging process can then become a candidate for automation. The construction industry is such an audience and each construction project provides an opportunity for shared perspective that enables a finite number of tags to have relevance for a bulk of the audience.  I look forward to exploring next leap in productivity gains that Photo-Data will make possible.

Advertisements

The Photo-Data Transformation

July 22, 2010

In my previous post I identified the cost associated with extracting information from the photos and video used to capture ground level information from construction project sites. Information extraction costs are a function of time once the project ends and eventually reach impractical levels as the project team disbands and their tribal knowledge of the project evaporates forever.

But it doesn’t have to be that way. What if the project team could encapsulate their tribal knowledge into an archive that remained accessible to the project team and the owner of the finished building for the duration of the building’s life? – Time out. You may be thinking, um, OK Geedra blog dude. Are we talking about some kind of Vulcan Mind Meld?

By packaging the visual media (photos and/or video) with the project team’s intimate knowledge of the images in the media into photo-data, we can make that information accessible without incurring the extraction costs associated with current forensic research efforts using construction photos and video. The components of photo-data include data tags (for indexing and research), annotations (tells the story of the image), contextual references (other photos, floor plans, reports) and any other supporting information.

(c) 2010 Geedra, LLC

Once converted, photo-data information will be resistant to degradation over time, making it as accessible 10 years after completion as it was 10 minutes after the photos were taken. Facility managers and building owners can not only benefit from knowing the history of their building’s past, but can continue to build upon their photo-database by adding new photos to capture conditions during building improvements and major maintenance overhauls. In doing so, they are making a major improvement to the value of the physical asset in a move that will lower operating costs and reduce operation risk for the life of the building.