After last saturday we managed to implement face detection inside the Julio video stream, with 2 more days working on face recognition, we have now put it to work. We can now load a photo onto Maria software and look for it on the Julio video capture.
We have now to run some tests with different environment conditions (lighthing, speed, accurancy with more different faces and types of photos, etc.) and produce some fine tunning to the system.
A major milestone was achieved in order to know if the project was viable. It's seems we are good to go!