MosaHIP: Mosaic-based Human Interactive Proof

Below are presented two prototypes of MosaHIP, a Human Interactive Proof (HIP), also known as CAPTCHA, based on mosaic of pictures. Its purpose is to discriminate between human actions and computers activities, in order to prevent massive download of web contents performed by automated bots.

On the left side of the page of each demo, there are some web resources (html pages, videos, documents, etc.) which are not shown as links, but as movable text strings. To donwload a specific resource, one has to drag and drop the associated string on the picture indicated by the system.
If the procedure is successfully accomplished, the resource is placed in the basket and it can be accessed by clicking on the Show button. Otherwise, after a number of wrong attemps, the system simply locks the user who keeps failing the test by preventing the his/her IP address from accessing the web site. The lock lasts for a period of time, which increases in length when the number of errors from the same IP address grows; then, the test is presented again.

Further information can be found here or by writing us an .

Two different tests are presented:

  • The concept-based MosaHIP, which is focused on identifying a specific concept. To pass the test, one has to click on the picture indicated by the HIP in red.

    For example, if the requested picture is scorpion, one should locate the picture showing the scorpion (indicated by the green arrow) within the set of images and click on it.

    Note that pictures may overlap each other and several of them display non-existent contents.


     
    Mosahip (concept-based)

     
  • The topmost MosaHIP, which is based on the top-most picture: to pass the test, you have to click on the top-most picture contained within the set of images.

    In the example on the right, the image to be selected is the the one showing a panda, pointed by the green arrow.

    Note that such a picture represents something existing and lays over the others, not being overlapped by other images.


     
    Mosahip (top-most)

     


 
Computer and Network Security Group, Dipartimento di Informatica - Università degli Studi di Torino