ESP32-Object-Recognition

My humble attempt at getting tensorflow and the ESP32-CAM to cooperate (among other things)

Notes:

"Webcam_detection.py" isn't plug and play. If you plan on running it, you probably know but you're doing but here's a couple of pointers:

• Get OpenCV & TensorFlow
• My project uses TensorFlow 1.0 which only supports Python <= 3.7
• Get a TensorFlow model [COCO]

Similarly, the ESP32 project isn't plug and play either:

• Open the arduino code in PlatformIO as that should automatically resolve any library dependencies.
• Change it accordingly
  • Change the ssid/password
  • Set the correct resistors used for the voltage divider
  • Set server port/local ip
• Power is supplied through a 4.2 lithium ion battery powered through the appropriate regulator

Web:

• Set ESP module's local IP
• Try to make the webpage (index.html) look nice??

Also, the ESP32 doesn't allow the usage of ADC1 (or was it 2?) while the wifi radio is (WiFI.begin() has been called. On the esp32, this basically means that the module can't receive analog input at all (since all of the ADC2 pins are used for the camera sensor). In order to monitor my battery's cuurent capacity, I had to solder a wire to gpio 33 (connected to the inbuilt red indicatir led on the board)

I probably missed stuff, idk