A bounding box is an optional parameter that can be returned with every recognized item to indicate the position of the matched reference image inside the query image. More specifically, a bounding box represents the four corners of the matched reference image projected into the query image.
The order of the corners is: top-left, top-right, bottom
For every recognized item we return a score representing similarity between the matched object from the query image and the most similar reference image from the item.
Scores are represented with values from 0 to 100. The more similar are the matched objects the higher the score. Scores can be also seen as a measure of confidence that the matched
Yes, we have put in place a cross-domain policy file giving permissions to handle data across multiple domains. Simply contact us indicating your domain, and whether you are using https or not, and we will enable it for you.
Yes, when using Flash, you can use our Image Recognition API, implementing functions that capture and send images to our service. Check out the Image Recognition API and the best practices for query images. Note that that you’ll need to ask support to provide cross-domain policy for you.
All query images are being saved. Currently you cannot access them via our APIs, but you can contact us to get them.
Our service performs the recognition within less than 500 ms.
The total recognition round-trip depends on your network bandwidth. Normally on Wi-Fi and 4G you get the results within 600-800 ms. On 3G the round-trip varies from 1 to 2 seconds.
A Token is a string of characters that uniquely identifies the collection of images that you want to search through. Tokens are used in the Image Recognition API, in the CraftAR Mobile App, and in the CraftAR mobile SDKs to specify that collection.
If you’re trying to find the default token for your collection, follow these steps:
Log in to the
Receiving the metadata associated with the recognized items is optional. A specific metadata field (e.g. custom data) can be included using a corresponding parameter in the recognition requests.
Visit our documentation on how to do it with our Image Recognition API or our mobile SDKs.