Abstract: A smart postal mailbox device is disclosed that incorporates a camera and an imaging processing capability. When the device is in place inside a postal mailbox and is activated by a content element being placed in, or removed from, the mailbox, the camera obtains an image of the interior of the mailbox, which it then analyzes by first using pre-processing to correct for geometric distortion. The image is then decomposed using vision related algorithms into a discreet information dataset. This discrete data set is compared to previous data sets using a neural decision tree. If appropriate, the device transmits a message informing a remote user of the current state of the mailbox.