loading
This Thread  New Post   Forum Index 

 
MS Tags
Bill from United States  [1 posts] 		16 year
Has anyone used RR to spot or decode microsoft "tags"? http://blogs.msdn.com/tag/

Thanks for the Denver Maker presentation, and congratulations on attracting Pi Robot whose tutorials offer amazing demonstrations of your work!
Anonymous 16 year
Bill,

We can spot and read them quite easily ... the issue is that once read they are meaningless unless you can interface with MS to actually get the decoded reference number. MS has patenting the tag/process and thus is unlikely to release how they specifically encode the tag and how it should be decoded for use.

We are however, working on other non-proprietary barcodes like QR codes.

What there a specific reason you are interested in MS tags?

STeven.
Anonymous 16 year
STeven -

I appreciate your thoughtful response.  Unfortunately I have no immediate application for the MS tags, but have been intrigued with their range of applications.

Your s/w holds the key to managing an app which could be independent of ms, aside from patent / engineering issues of course.

They use an array of 10 by 5 spots in a fixed field; each must hold one of four colors.  If the field can be located and the spots read, then the encoding for the tag is 2 bits/spot * 50 spots.  One could maintain her own database of tags created (with a tag-builder) and their assigned significance.  The tag-builder might be a web service, with a printing web service as well.  A significance posting could be a web service, and certainly retrieving the significance from a reading is a fourth web service component.  It would appear this is a whole class of apps, of a variety of fixed field shapes and arrays of visual code spots, and could be applied outside the general marketplace attracting microsoft.

Of course, the principle of recognizing a spot pattern in each painting in a museum or home (pi robot mentioned this, in terms of color histograms?) uses this technique to define reference points with no intrusion of marking up the environment.  The ms tags do carry the information capacity to discriminate between 2^100 objects;  who needs that many?  

What I have seen tells me that this approach can trivialize some recognition problems, a huge flexibility with a small amount of technology.

Obviously this is in the same league as the QR codes.  The approach in general provides a way to balance speed and reliability of reading with the information content required per tag and the environmental constraints.  The basic QR codes can carry 30 times as much data (well, 30 times as many bits) as the ms tag.  That could deliver rules for handling the labeled object where a simple web access is not available to expand a code.

I've posted some sample ms tags.  One precisely showing the spots, one with triangles filling the space.  One only using black dots.  And then the sample QR code from its wikipedia.

Sorry I don't have an app at hand to build and deploy!

  ~  ~  Bill

 

This forum thread has been closed due to inactivity (more than 4 months) or number of replies (more than 50 messages). Please start a New Post and enter a new forum thread with the appropriate title.

 New Post   Forum Index