Jump to content

  • Log in with Facebook Log in with Twitter Log In with Google      Sign In   
  • Create Account

Subscribe to HRA Now!

 



Are you a Google Analytics enthusiast?

Share and download Custom Google Analytics Reports, dashboards and advanced segments--for FREE! 

 



 

 www.CustomReportSharing.com 

From the folks who brought you High Rankings!



Photo
- - - - -

Does The Google Spider Read Text In Images


  • Please log in to reply
11 replies to this topic

#1 dibbern

dibbern

    HR 2

  • Members
  • PipPip
  • 10 posts
  • Location:Chicago

Posted 29 July 2013 - 07:11 PM

I know this is a topic that comes up every year or so. The answer has been, I believe, no, the G crawler does not read text in images.

 

But I'm wondering now, in light of patents filed for G to read text in images, and news that G translator can translate text in images. Can G read those boiler-plate gifs many of has used for years to demote repeated text?

 

 



#2 qwerty

qwerty

    HR 10

  • Moderator
  • 8,628 posts
  • Location:Somerville, MA

Posted 29 July 2013 - 07:40 PM

They're getting there, but I don't think they've arrived just yet. You should still rely on the image's alt attribute.

 

As I understand, Google has officially announced that they've managed to identify a picture of a cat as a picture of a cat. I'm not aware of anything beyond that. You would think that text would be relatively easy, though.



#3 dibbern

dibbern

    HR 2

  • Members
  • PipPip
  • 10 posts
  • Location:Chicago

Posted 29 July 2013 - 08:48 PM

If it was really hard, we wouldn't have all these darn captchas, would we? 



#4 bobmeetin

bobmeetin

    HR 6

  • Active Members
  • PipPipPipPipPipPip
  • 535 posts
  • Location:Colorado

Posted 29 July 2013 - 08:49 PM

Regarding the image itself, I'm sure Jill is on target.  Easy to do - I don't know. They would need to employ some type of OCR software which would present an educated guess at best and take some development cycles. But why?

 

However, images contain hidden information about the picture, it is called EXIF, you can google exif. When you take a picture some technical data is recorded and stored with the image.  Besides the technical data you can also edit/update and store some picture information such as caption, description, keywords, author, and geo-data (geotagging). Picasso, flickr and some other apps give you some access to update this info. Along with that come multiple ways to read this text including the browser image properties and the properties seen when you right-click on an image saved to your computer. 

 

Google maps/plus are supposedly reading some of this, including the geotagging. I suspect that this will grow over time, until it becomes a thing.



#5 dibbern

dibbern

    HR 2

  • Members
  • PipPip
  • 10 posts
  • Location:Chicago

Posted 29 July 2013 - 10:18 PM

My concern was with a few boilerplate text elements that I converted to gifs years ago to keep from repeating content on every page. I'm wondering now: is G reading those, and factoring in the repeat content as a negative?



#6 bobmeetin

bobmeetin

    HR 6

  • Active Members
  • PipPipPipPipPipPip
  • 535 posts
  • Location:Colorado

Posted 29 July 2013 - 10:29 PM

Odds are, 99.9999% no. You can always go to google, look for this exact text, quoted, in context of your website. Google is not dumb, occasionally a difficult child, but not dumb.



#7 qwerty

qwerty

    HR 10

  • Moderator
  • 8,628 posts
  • Location:Somerville, MA

Posted 30 July 2013 - 05:09 AM

You know, it's OK to repeat content on every page, as long as every page has a decent amount of unique content to set it apart from the rest. Google isn't going to slap you for having something like "Copyright 2013 Company X, Inc." on every page. Even if you stick a paragraph of legalese in small print on every page, you're fine.

 

Just what sort of boilerplate have you been replacing with images? And if you're hoping that Google is now able to read the text in the images, wouldn't that be the same thing as having them simply read the text if you hadn't replaced it with gif files?


Edited by qwerty, 30 July 2013 - 05:10 AM.


#8 Jill

Jill

    Recovering SEO

  • Admin
  • 33,011 posts

Posted 30 July 2013 - 07:08 AM

My concern was with a few boilerplate text elements that I converted to gifs years ago to keep from repeating content on every page. I'm wondering now: is G reading those, and factoring in the repeat content as a negative?

 

It is absolutely not a problem. First, they don't read it, but even if they did, they don't care as it's part of your template. 

 

Do you change your navigation on every page so that it won't be the same? How about your logo and tag line?  I sure hope you don't as that would be pretty silly. 



#9 Mikl

Mikl

    HR 5

  • Active Members
  • PipPipPipPipPip
  • 310 posts
  • Location:Edinburgh, Scotland

Posted 30 July 2013 - 07:13 AM

When it comes to "reading" images, Google is much cleverer than many of us might think. To demonstrate, take a look at their reverse image search tool. Show it, say, a photo of someone cycling along a beach. It will show you more pictures of people cycling on beaches. These are not simply cropped or resized versions of the original. Nor are they photos that have the same Alt attribute as the original. They are completely different images that happen to depict the same subject.

 

On that basis, being able to read text within an image would seem fairly trivial. It's almost certain that Google can do that (but whether they have a commercial reason for doing so is another mattter).

 

Mike



#10 imexperttools

imexperttools

    HR 2

  • Members
  • PipPip
  • 29 posts
  • Location:Frankfurt

Posted 30 July 2013 - 08:25 AM

I believe Google has gone a long way in the field of image recognition all thanks to their enormous database of information which they have been storing over the years. Google glasses is a prime example of how far Google has gone with the image recognition technology. Get the image and compare it with their organized database in order to generally understand the idea of the image is not that difficult.

 

If they can do that then i think the text in images is not that big of a challenge for them. You can expect in another year or two, the content within video, images and even documents would be something that you would have to consider as SEO factors.



#11 qwerty

qwerty

    HR 10

  • Moderator
  • 8,628 posts
  • Location:Somerville, MA

Posted 30 July 2013 - 11:15 AM

Google glasses is a prime example of how far Google has gone with the image recognition technology. Get the image and compare it with their organized database in order to generally understand the idea of the image is not that difficult.

I think you mean Google Goggles, which is not the same thing as Google Glass. I took some pictures on my phone a few days ago, and noticed a little while later that Goggles was running -- I hadn't opened the app. It had identified a picture on the wall of a room in which I had taken a picture.

 

But again, that's comparing the form to other images in their database. It's not the same thing as interpreting the form to analyze its meaning. If you show Google a picture of a person crying, Google can probably find other pictures of that person, and it can probably find other pictures of people crying, but I don't think it can tell it's a picture of someone who's sad unless it or some of the pictures it considers similar are on pages containing text that makes that point.



#12 bobmeetin

bobmeetin

    HR 6

  • Active Members
  • PipPipPipPipPipPip
  • 535 posts
  • Location:Colorado

Posted 30 July 2013 - 12:39 PM

Just because they can, doesn't mean they should. They need a strong business reason to break old-school SEO indexing and SEO methods. Somewhere you draw the line.

 

The fact that they're massaging page titles (and descriptions) on a case by case basis is already a nuisance. Seems to me that some things should be left to the business owner or SEO delegate to manage without attempting to reinterpret on their behalf leading to potentially puzzling results. Images would be a further step beyond the line and over the edge.






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

SPAM FREE FORUM!
 
If you are just registering to spam,
don't bother. You will be wasting your
time as your spam will never see the
light of day!