Try to follow these steps 1 - Tenserflow and OpenCV installation (A) 2 - Make sure your Camera is working 3 - Install requirements.txt 4 - Face recognition library (B) A: For Tenserflow and OpenCV installation and problems checkout this video Tenserflow and Opencv installation for Pi Repo for matching Python-Tenserflow on a Raspberry Pi- architecture B: For Face recognition library checkout this video Face recognition library
Credit and Thanks to Electrical Engineer Hussain Balhareth He did the bar training and he gave me the chance to work on this project, he also did troubleshooting and testing for libraries and hardware issues
#-------Raspberry pi --------- # Common # pip==22.0.4 six==1.16.0 numpy==1.22.3 # CV2 Most already installed from "Installtion Notes" section Above imutils ==0.5.3 #face-recognition ==1.3.0 #face-recognition-models ==0.3.0 #tensorflow ==2.7.0 #tensorflow-estimator ==2.7.0 #keras ==2.7.0 #Keras-Preprocessing ==1.1.2 # opencv-contrib-python ==4.5.5.64 # opencv-python ==4.5.5.64 # dlib ==19.23.1 # Probably needed if you're using the Pi Camera -> Remember usb Cameras are easier to deal with # picamera == 1.13 -> If this doesn't work - Probably won't it's old and not for new Pis -> OsError libmmal.so # Consider using PiCamera2- Currently you'll have to build it though # Not Sure scipy == 1.8.0 matplotlib ==3.5.1 # ------- Others ----------- # Pillow == 9.1.0 # flatbuffers ==2.0 # Probably needed -> argparse # For keyboard stuff -> pynput #------------------- Pi end----------------------------