Security webcam running open-cv for object detection. The camera will send an email with an image of any objects it detects. It also runs a server that provides a live video stream over the internet.
To get emails when objects are detected, you'll need to make a couple modifications to the mail.py file.
Open mail.py with vim vim mail.py, then press i to edit. Scroll down to the following section
# Email you want to send the update from (only works with gmail)
fromEmail = '[email protected]'
fromEmailPassword = 'password1234'
# Email you want to send the update to
toEmail = '[email protected]'
and replace with your own email/credentials. The mail.py file logs into a gmail SMTP server and sends an email with an image of the object detected by the security camera.
Press esc then ZZ to save and exit.
You can also modify the main.py file to change some other properties.
email_update_interval = 600 # sends an email only once in this time interval
video_camera = VideoCamera(flip=True) # creates a camera object, flip vertically
object_classifier = cv2.CascadeClassifier("models/fullbody_recognition_model.xml") # an opencv classifier
Notably, you can use a different object detector by changing the path "models/fullbody_recognition_model.xml" in object_classifier = cv2.CascadeClassifier("models/fullbody_recognition_model.xml").
to a new model in the models directory.
facial_recognition_model.xml
fullbody_recognition_model.xml
upperbody_recognition_model.xml
Run the program
python main.py
You can view a live stream by visiting the ip address of your pi in a browser on the same network. You can find the ip address of your Raspberry Pi by typing ifconfig in the terminal and looking for the inet address.