The example code is under git repository: a311d/ai_examples/face_recog_callback. It realizes the functions as below:
1) Detect the human face in the video
2) Match the detected face against the pictures in the facial library
3) If matching, the registered callback would be called for jpeg picture data and AI information reporting
Step 1: Edit the code and cross compile it
cd a311d/ai_examples/face_recog_callback/src
make
scp the executable binary to the boardStep 2: Build up the facial picture repository
scp the picture to the board
testaipiclab -l telib -t 0 /*create lib*/
testaipiclab -l telib -t 1 /*delete lib*/
testaipiclab -l telib -t 2 -i 000 -n zcj -p /var/piclab/tmp/000_zcj.jpg /*add pic*/
testaipiclab -l telib -t 3 -i 000 -n zcj /*del pic*/Step 3: Run the program
The AI parsed information would be printed out on the console whenever the matching face is detected.
-
The AI algorithm has the temporary license, so it could only be used for 2-3 hours
-
The AI algorithm has not been fine tuned, so the accuracy is not so good.
-
If you want the commercial AI algorithm, please contact us with email: [email protected]