问题
I have a collection of profile images from customers I need to be able to pass a selfie of the person and scan it across the collection of images and pull up the customer information.
Need to do the following using AWS Rekognition -
- Create a collection - Done
- Add Images to the collection - Whats the REST API syntax for this
- While adding the images to the collection also tag it with the customer name.
- Take a selfie portrait and search across the collection and return the tag information which matches.
Im using Flutter as a platform hence there is no support for AWS SDK so will need to make REST API calls. However the AWS docs don't provide much information for REST support.
回答1:
The APIs are documented. For example to detect faces in an image and add them to a collection, see IndexFaces.
I'd personally recommend getting comfortable with Rekognition via the awscli (or Python/boto3) briefly before you move to the Rest API.
On the name tagging front, you assign an 'external ID' to faces when adding them to a collection. That external ID is the correlator that you supply and that Rekognition stores. Later, when you ask Rekognition if a given face matches one already in a collection, Rekognition will return you the external ID. That can then be used as a lookup into some database that you have to identify the person's name, date of birth, or whatever.
来源：https://stackoverflow.com/questions/55603140/need-to-compare-a-given-face-input-image-across-an-already-stored-labelled-colle