Cloudinary Blog

How to automatically tag and categorize photos according to their content

How to automatically tag and categorize photos by their content

Update - December 2015: The add-on described in this post is no longer available since ReKognition terminated their services. However, all features described here are still available via a different and even better add-on by Imagga. See Automatic image categorization and tagging with Imagga and the Imagga Auto Tagging Add-on documentation.

Drawing insights from user generated content can be very useful. If you allow users to upload images, you might want to better understand what their images contain. Whether a photo is of a landscape, people, animals, or nightlife, image processing and analysis can assist in further comprehension.

ReKognition scene categorization

Cloudinary takes care of the entire image management pipeline, from uploading, via dynamic image manipulation, to fast CDN delivery. You can use Cloudinary with our add-on for automatic scene categorization provided by ReKognition, a visual recognition solution developed by Orbe.us.

ReKognition logo The ReKognition scene categorization add-on analyzes scenes within photos and allows you to automatically classify your website’s images into a long list of potential categories.

Improve user experience, engagement and conversion

By automatically recognizing what images contain, developers can easily create pages or site sections based on images’ content (e.g. photos of animals, travel, special events, etc.). In addition, if your site supports a search option, you can have the search results include relevant images based on their visual content instead of just by their file names or user-assigned tags. These benefits can improve user experience on your site and increase engagement.

When you identify image content, you then know what images your users upload or search for and can target them with related content in order to enhance engagement and conversion. For example, if a user uploaded or searched for cat images, you can target him/her with actions, such as presenting pictures of other pets or any cats’ related content. Knowing the content of images on your site can have a direct impact on your online business.

How-to automatically categorize your images

Web and mobile developers can use Cloudinary's API to process images while uploading or process previously uploaded images.

For example, the image below was uploaded to Cloudinary using the code below. You can see that the image’s content is clearly of a dance party, so let’s see how it is categorized with the ReKognition add-on. Simply set the categorization upload parameter to rekognition_scene in order to perform the image’s content analysis and get the detected categories:

Cloudinary::Uploader.upload("my_dance_photo.jpg",
          :categorization => 'rekognition_scene', 
          :public_id => "dance_party")

Ruby:
cl_image_tag("dance_party.jpg")
PHP:
cl_image_tag("dance_party.jpg")
Python:
CloudinaryImage("dance_party.jpg").image()
Node.js:
cloudinary.image("dance_party.jpg")
Java:
cloudinary.url().imageTag("dance_party.jpg");
JS:
cloudinary.imageTag('dance_party.jpg').toHtml();
jQuery:
$.cloudinary.image("dance_party.jpg")
React:
<Image publicId="dance_party.jpg" >

</Image>
Angular:
<cl-image public-id="dance_party.jpg" >

</cl-image>
.Net:
cloudinary.Api.UrlImgUp.BuildImageTag("dance_party.jpg")
Android:
MediaManager.get().url().generate("dance_party.jpg");
iOS:
imageView.cldSetImage(cloudinary.createUrl().generate("dance_party.jpg")!, cloudinary: cloudinary)
Sample dance party uploaded photo

Cloudinary’s upload API's result shows the uploaded image's detected categories as well as their level of accuracy, provided as a score between 0 and 1, where 1 is 100% accurate. As you can see in the example below, the ReKognition add-on detected several categories for the image above, that were each given a score. The best category match was 'dance', with a 0.719 accuracy score.  

    {"public_id"=>"dance_party",
     "url"=>
      "https://res.cloudinary.com/demo/image/upload/v1417446163/dance_party.jpg",
     ...
     "info"=>
      {"categorization"=>
        {"rekognition_scene"=>
          {"status"=>"complete",
           "data"=>
            {"matches"=>
              [{"tag"=>"dance", "score"=>0.719},
               {"tag"=>"blonde", "score"=>0.3055},
               {"tag"=>"model", "score"=>0.24},
               {"tag"=>"night_club", "score"=>0.1859},
               {"tag"=>"hair", "score"=>0.1449}]}}}},
    }

Image categorization data can be stored on your side in order to organize your content and target user interest. You can then take actions such as linking the categorization data to users that upload categorically related images. Once the information is captured, a data analysis can take place generating insights, including users’ image preferences. Based on the information and analysis provided, you will be able to take proactive actions to enhance user engagement and conversion.

Automatic tagging with ReKognition scene categorization

Cloudinary allows you to tag uploaded images. Each image can be assigned one or more tags. This feature automatically assigns tags based on categories that are detected by the ReKognition add-on.

You can see below that we set the auto_tagging parameter to 0.15 for the dance party example image above, which means that auto tagging will use the detected categories above this threshold. You can set the threshold to a higher level if you want to be more strict regarding the detected categories you want to tag. As for our example, the tagging results were 'blonde', 'dance', 'model', and 'nightclub':

Cloudinary::Uploader.upload("my_dance_photo.jpg", 
      :categorization => 'rekognition_scene', 
      :auto_tagging => 0.15, 
      :public_id => "dance_party")

Results in:

    "tags" => ["blonde", "dance", "model", "night_club"]

With automatic tagging using ReKognition’s scene categorization add-on, you will be able to use Cloudinary's admin API in order to list and search for all images with specific tags.

Summary

Understanding user generated content can provide you with great insights, allow you to take actions such as creating dynamic content pages, and match content to user preferences. Cloudinary’s service together with the fully integrated ReKognition scene categorization add-on provides developers with the powerful ability to enhance their image content management as well as increase their online users’ engagement and conversion.

ReKognition add-on

The ReKognition scene categorization add-on is available with all Cloudinary plans, You can try it out with the add-on’s free tier. If you don't have a Cloudinary account yet, sign up for a free account here.

Update - December 2015: The add-on described in this post is no longer available since ReKognition terminated their services. However, all features described here are still available via a different and even better add-on by Imagga. See Automatic image categorization and tagging with Imagga and the Imagga Auto Tagging Add-on documentation.

Recent Blog Posts

Hipcamp Optimizes Images and Improves Page Load Times With Cloudinary

When creating a website that allows campers to discover great destinations, Hipcamp put a strong emphasis on featuring high-quality images that showcased the list of beautiful locations, regardless of whether users accessed the site on a desktop, tablet, or phone. Since 2015, Hipcamp has relied on Cloudinary’s image management solution to automate cropping and image optimization, enabling instant public delivery of photos, automatic tagging based on content recognition, and faster loading of webpages. In addition, Hipcamp was able to maintain the high standards it holds for the look and feel of its website.

Read more
New Image File Format: FUIF: Why Do We Need a New Image Format

In my last post, I introduced FUIF, a new, free, and universal image format I’ve created. In this post and other follow-up pieces, I will explain the why, what, and how of FUIF.

Even though JPEG is still the most widely-used image file format on the web, it has limitations, especially the subset of the format that has been implemented in browsers and that has, therefore, become the de facto standard. Because JPEG has a relatively verbose header, it cannot be used (at least not as is) for low-quality image placeholders (LQIP), for which you need a budget of a few hundred bytes. JPEG cannot encode alpha channels (transparency); it is restricted to 8 bits per channel; and its entropy coding is no longer state of the art. Also, JPEG is not fully “responsive by design.” There is no easy way to find a file’s truncation offsets and it is limited to a 1:8 downscale (the DC coefficients). If you want to use the same file for an 8K UHD display (7,680 pixels wide) and for a smart watch (320 pixels wide), 1:8 is not enough. And finally, JPEG does not work well with nonphotographic images and cannot do fully lossless compression.

Read more
 New Image File Format: FUIF:Lossy, Lossless, and Free

I've been working to create a new image format, which I'm calling FUIF, or Free Universal Image Format. That’s a rather pretentious name, I know. But I couldn’t call it the Free Lossy Image Format (FLIF) because that acronym is not available any more (see below) and FUIF can do lossless, too, so it wouldn’t be accurate either.

Read more
Optimizing Video Streaming and Delivery: Q&A with Doug Sillars

Doug Sillars, a digital nomad and a freelance mobile-performance expert, answers questions about video streaming and delivery, website optimization, and more.

Doug Sillars, a freelance mobile-performance expert and developer advocate, is a Google Developer Expert and the author of O’Reilly’s High Performance Android Apps. Given his extensive travels across the globe—from the UK to Siberia—with his wife, kids, and 11-year-old dog, Max, he has been referred to as a “digital nomad.” So far in 2018, Doug has spoken at more than 75 meetups and conferences!

Read more