Recently google launched its new feature in image search by image means we can search other images by uploading a image in the google search box. How is this possible?
-
What aspect are you asking about? – Oded Jun 20 '11 at 10:59
-
we can search images by dropping images in search box......how is works..... – Kumod Jun 20 '11 at 11:02
-
if you are interested there is this related site: [Tineye](http://www.tineye.com/about) which does reverse image searches – jemmanuel Jun 20 '11 at 14:52
3 Answers
Look at WP:Content-based image retrieval. An example of open-source implementation that you can study internal working of is for example GNU Image Finding Tool.

- 73,652
- 13
- 125
- 172
If you click on the "Learn more" link on the page you are referring to, you'll find this explanation
How it works
Google uses computer vision techniques to match your image to other images in the Google Images index and additional image collections. From those matches, we try to generate an accurate "best guess" text description of your image, as well as find other images that have the same content as your search image. Your search results page can show results for that text description as well as related images.

- 1
- 1

- 44,604
- 7
- 83
- 130
Actually the answer to this lies in the image processing.....in over a decade image processing and computer vision have done great deal of advancement... search by image uses pixels ...it compares the pixels and matches with image database it contains....its quite similar to what actual tyext search does but there pixels in place of text... there are various operators like soble operator,etc which help us focus on the important details of the picture being tested and and we we can search on the basis of the important feature of the image.....

- 506
- 1
- 4
- 9
-
It doesn't search by pixels nor with sobel like operator, as far as I know, google uses scale invariant features and tinyeye uses variation of MSER – om-nom-nom Jun 19 '12 at 20:36