Multiple object recognition now available in our API

Some time ago, we wrote about multiple object recognition from one image. Now it is available in our API! 

You can imagine a situation in which a user takes a photo and there are two or more products you would like to recognize in his or her photo. Let’s consider the following photo taken by a user:

 

With Recognize.im Pro Account, it is possible to recognize multiple objects in this picture. To use this feature, you need to follow these four easy steps:

 
1. Upload your reference photos to our service. Let’s say we upload the photo of the tea box and basil spice. Here is what our Manage Photo page looks like:

2. Click Apply button:

3. We need to change the recognition mode to multi. To do this, click Change mode to multi button:

 

This can also be done from your code by calling the modeChange SOAP method.

4. In your application call http://recognize.im/recognize/multi/xxxx where xxxx is your Client ID.
In the answer, you will get a JSON response with the list of IDs of images that were recognized in the photo from the query. In our case, the response is the following:
{“status“:0,”id“:["tea","basil"]}

Take a look at our documentation for more information. You can also check out our sample codes for iOS and Android which show exemplary usage of multiple objects recognition. You can also visit our Github repository where examples for different programming languages are stored.

 

Author:
Jakub Jurkiewicz
Recognize.im Developer

  • theguy126

    Why is there not yet a shopping app where I can take a picture of a whole aisle of beer and sort by their rating on beeradvocate? Let me guess: Maybe it’s because the API’s that actually work are priced so high that no developer wants to bite the bullet and take the risk of paying a bajillion dollars per year for the 1% chance that the much-needed app actually takes off virally.

WordPress SEO fine-tune by Meta SEO Pack from Poradnik Webmastera