Which texts does the robot use for image search?

The alt attribute is an alternative source of information for users who have disabled images in their browser. If the alt attribute is specified, the attribute text will display when it's not possible to display the image:

If the alt attribute is not specified, the image space will be empty:

It's especially important to describe the image by specifying the alt attribute for sites whose content consists predominantly of images.

This is how an alt attribute should look:

<img border="0" alt="[alternate text]" src=​"risunok.jpg"
width="[image width]" height="[image height]">

You don't need to put all the site's or page's keywords in the alt attribute, only the ones that really describe the image. This will help the robot find the most appropriate image, and make it easier for the Yandex user to select it from the search results and get to your site.

The title attribute gives additional information about the image. The text of this attribute appears when you hover your mouse over the image:

This is how the title attribute should look:

<img border="0" alt="[alternate text]" title="[caption]"
src=​"risunok.jpg" width="[image width]" height="[image height]">

If the image is a link, then it's best to put the title attribute in the link element:

<a href="link.html" title="[caption]"><img border="0" alt="[alternate text]"
src=​"risunok.jpg" width="[image width]" height="[image height]"></a>

In addition to the alt and title attribute values, the following texts in image searches:

• The texts of links to images from other pages and sites

• The text on the webpage that is located in close proximity to the image

• Texts and headers of short documents that surround a single image

• Names of files and image scripts (including their transliterations and simplified word-for-word translation).