I am currently developing a computer vision project based on Python, OpenCV, Raspberry pi, firebase and android.
I am using face detection, recognition and tracking approach in the project but I am facing integration issues and need code optimisation. Most of the components of the project are ready I just need someone to help me integrate it according to the required architecture. All the details and existing the code will be provided
I am using webcam for capturing the image.
The webcam is connected to a raspberry pi.
I have done face and upper body detection using Haar cascade.
Facial tracking with Kalman filter
Facial recognition with dlib and face recognition library.
I have also developed the android app and all the databases on firebase.
I intend to do detection and tracking on the pi and recognition on the server. Where everything is connected to the firebase I am not able to transfer the images from real-time stream to server along with its tracking ID as well as retrieve the recognised person's name and id from the server
I even need optimisation for the tracker and optimising the code on raspberry pi because FPS is really low
The camera will be installed above the door and I intend to send a notification on the android app if the recognised person leaves the room