This Python library allows you to automatically record ip-cam videos on your Google Drive storage. You are allowed to record videos from an arbitrary number of cams you might have installed on your local network. The records are performed on the RSTP protocol.

The application has the intention to allow the user to be safer at home, being able to check recorded video remotely in a near-real time fashion.

How it works

The application uses ffmpeg to record videos from ip-cams connected to the local network.
The records are saved locally on the /video path and then automatically loaded on Google Drive. If the storage amount
occupied by the videos exceeds a certain threshold, oldest video are identified and erased from the Google Drive storage. The same occurs if the local video storage exceeds a certain (and by default much larger) threshold. This allows to continuosly running the application without worrying of storage issues.

Install

Install the cams

The first step is to have installed the ip-cams you want to use on your local network. Depending on the cam, you might need to perform different operations. Please follow your cam vendor's instruction to install your cams.
After installing, you should be able to access to your cam connecting to the local ip address on the port RSTP. For example, if you have installed your cam at IP 192.168.1.44 on port 554 (the usual port used for handling RSTP connection), you should be able to open your browser and connect your cam when inserting 192.168.1.44:554 on the search bar.

Set up Google Drive API access

The second step is to allow the program to interact with the Google Drive API. In order to do that, you need to have a valid Google Cloud account. Then, you need to create a new project and set up the credentials. Finally you can enable the Google Drive API on the project.

• to create a project: here
• To set up the credentials: here. When you are setting up the service account, you are required to download a json file containing the private key which allows to connect to the google cloud project. The downloaded json must be named credentials.json and be placed in the path ./configs/credentials/credentials.json.
• To enable the Google Drive API: here

Create storage folders on Google Drive

Next, you need to create the folders you want to store data on Google Drive. To do that, it is enough to connect to google drive and create the folders you want to use to store your recorded videos. You might want to create a different folder for each cam you want to record from. When creating the folder, take note of the folder id. This can be retrieved by looking at the last part of the url you see when entering in the google drive folder. For example, in the following url

https://drive.google.com/drive/folders/78643Yd_YjwDoFnPIbzdI82ACT0K

the folder id is 78643Yd_YjwDoFnPIbzdI82ACT0K

Configure the cam details

You need to tell the application which is the local ip address used by the cams, the RSTP port, the password and username which can be used to connect to the cam. You need to create a json file storing this information on the path ./configs/cams. Each cam has to have its json file saved in this path. The name of the json file is not important. The json file has to have the following structure:

{
  "name": "30",
  "url": "192.168.1.30",
  "port": "554",
  "user": "admin_12345",
  "pwd": "54321",
  "folder_id": "123DvMivY_jUDoIUYNRPOLKQQbDQ"
}

This will tell to the program that there is a cam which we name 30 on the url 192.168.1.30 and port 554. To access the cam the user is admin_12345 and the password is 54321. The videos should be saved on the google drive folder having id 123DvMivY_jUDoIUYNRPOLKQQbDQ

Set up the parameters

The application can be customed according to different parameters. These can all be specified by modifying the YAML file contained in the path config/confs.yaml. The following is a list of the parameters which can be changed, reported with their default values:

• check_bdw : specifies if the application should turn-off when the internet bandwidth is not large enough. By default, when the cam is recording this option is disabled (False). Set to true to enable it.
• bdw_threshold : if the current internet bandwidth is below this threshold and check_bdw is set to True, the recording is not performed. In this case, the application sleeps for sleep_time_not_high_bdw. Instead, if this is not the case, the next bandwidth evaluation is performed after sleep_time_high_bdw seconds. By default the threshold is 4Mbit and the sleep times are respectively 300 and 60 seconds.
• allowed_schedule : specifies the allowed times in which the recording are performed and the local and google drive storing are enabled. It is a dictionary which has the weekday (0 = Monday, .., 6 = Sunday) as key and the interval times as allowed values for recording. If the current time is not included in the allowed time range, the next evaluation to verify it the current time is allowed is performed after sleep_time_not_allowed_time seconds. By default the allowed time ranges are every day 0-10am and 10-12pm. The sleep time is set to 60 seconds
• ext : specifies the file extension to be used by ffmpeg. By default it is set to mp4
• dimMB : specifies the dimension of each recorded video in megabytes. By default it is set to 1 megabyte.
• compression : specifies the video compression. By default is is set to 28.
• localLim: specifies the local storage threshold after which oldest videos are removed. By default it is set to 1 gigabytes.
• driveLim: specifies the drive storage threshold after which oldest videos are removed. By default it is set to 500 megabytes.

Execute

The program can be easily executed on docker. Once you installed docker it is enough to run:

docker-compose up