Auther Embedded
Device usage
During the launch, the device has to be connected to the internet to check its license.
- The whole face is seen by a camera.
- Being motionless within 0.15 sec.
- The distance is less than 50 - 70 cm ±5 cm depends on the settings.
- Face plane is parallel to the plane of the camera. The angle between their normals should not exceed 17 ± 3 degrees.
3D Scanner license activation
API is used for faces` database management along with facial recognition.
1. Register on the https://auther.com website.
2. Log into your Account.
3. Go to the License section and press "Add license" button.
4. Choose "I have activation code", and provide your activation code
5. Special installation command is generated now. Copy software installation command from the license line, which will be later used for software installation on the Up-Board single-board computer.
License grants
Depending on the grants which were added to the license on License tab, single-board computer may fulfill the following requests in Facecard Cloud:
- MANAGEMENT - faces' database management. Requests: /person/face/add, /person/face/update, /person/face/delete.
- IDENTIFICATION - facial recognition. Request: /person/face/recognize
Software installation on the UP-board
API is used for faces` database management along with facial recognition.
- Download and install Ubuntu 18.04.4 LTS (Bionic Beaver) 64-bit PC (AMD64)
- Connect Intel RealSense camera to the Up-Board computer for making device ready to use.
- Sign into Account , go to the license section and copy software’s installation command from the license line, assigned to the particular device.
- Paste software’s installation command into Ubuntu command line.
- Run the software’s installation command. During the process you need to create a Linux user password.
- After the installation has completed, the device is ready to start working.
- Default distance for face detection - 0.7m
Software configuration
Only two parameters can be managed:
- Maximum distance for face detection 0.5, 0.6, 0.7m
- Enabling the local recognition, the software remembers the last 5 unique faces and assigns them unique "faceKeys" that will be returned with a detected face image.
API description
Below you will find a description of working with the local API of the device. Communication with the local API occurs through communication via the local HTTP protocol.
Parameters description
- faceKey - temporary unique face identifier
- facePhoto - base64 encoded photo
- person_Id - url encoded identifier of the person
- code - error message
- message - human readable error description
Image description
Below are the requirements for the image that sent to our API. Also, with the help of the request "Get face image" you can get an automatically prepared image taken from the camera, which will meet the requirements below.
Image requirements
For a successful preparation of the photo, before the update, touch the offensive crocs
- Prepare an image with a custom face in JPG format
- Face size on the image should be more than 80 pixels
- Code the image for additional Base64. Received value is guilty of the offensive parameters:
- Image should be less than 300 kb
- Base64 characters count less than 400 000
Image verification criteria
To successfully create a customer ID, the image of the person that you send for registration will be checked for compliance with the following criteria:
- One face on the image
- Face position
- Face blurity
- Face size
- Face lighting check
- The similar face doesn't exist in the DB
Get face image
Receive the face which is in front of the camera or was no longer than 1 sec ago. As a result, you will receive an image that will meet the requirement above.
Request body
Response body
http status 200
{
“faceKey”:”17362cd0c”,
“facePhoto”:Base64format_string”
}
http status 400
{
“code”:”R_400”,
“message”:“Сamera is not connected”
}
Codes and messages for response 400
Create customer
Adding the face into the database
Request body
{
“personId”:”000391”,
“facePhoto”:”base64format_string”
}
Response body
http status 200
http status 400
{
“code”:”F_012”,
“message”:”Face exists”
}
Codes and messages for response 400
Update customer
Replace face in the database
Request body
{
“personId”:”000391”,
“facePhoto”:”base64format_string”
}
Response body
http status 200
http status 400
{
“code”:”F_012”,
“message”:”Face exists”
}
Codes and messages for response 400
Delete customer
Delete the face
Request body
{
“personId”:”000391”
}
Response body
http status 200
http status 400
{
“code”:”R_002”,
“message”:”Request body exception”
}
Codes and messages for response 400
Identify customer
Facial recognition in the cloud. If the “faceKey” parameter isn’t null, then the parameter facePhoto in that request is ignored and can be null.
The request /person/face/recognize could be submitted only with the last “faceKey” parameter, which was received in face/get request.
In that case the video module stores the last received image.
After the /person/face/recognize is submitted - the image is deleted from the memory.
Request body
{
“faceKey”:”17362cd0c”,
“facePhoto”:”base64format_string”
}
Response body
http status 200
{
“personId”:”000391”
}
http status 400
{
“code”:”R_002”,
“message”:”Request body exception”
}
Codes and messages for response 400
