Detect open RTSP hosts on any accessible subnetwork
Get their public info (hostname, port, camera model, etc.)
Bruteforce your way into them to get their stream route (for example /live.sdp)
Bruteforce your way into them to get the username and password of the cameras
Generate thumbnails from them to check if the streams are valid and to have a quick preview of their content
Try to create a Gstreamer pipeline to check if they are properly encoded
Print a summary of all the informations Cameradar could get

And all of this in a single command-line.

Of course, you can also call for individual tasks if you plug in a Database to Cameradar using the MySQL cache manager for example. You can create your own cache manager by following the simple example of the dumb cache manager.

Table of content

Docker Image
Quick install
- Dependencies
- Five steps guide
Manual installation
- Dependencies
- Steps
Advanced docker deployment
- Dependencies
- Deploy a custom version of Cameradar
Configuration
Output
Check camera access
Command line options
Under the hood
Contribution
Next improvements
Frequently Asked Questions
License

Docker Image

This is the fastest and simplest way to use Cameradar. To do this you will just need docker on your machine.

Run

docker run  \
            -v /tmp/thumbs/:/tmp/thumbs \
            -e CACHE_MANAGER=your_manager \
            -e CAMERAS_PORTS=your_ports \
            -e CAMERAS_SUBNETWORKS=your_subnetwork \
            ullaakut/cameradar:tag

your_subnetwork can be a subnet (e.g.: 172.16.100.0/24) or even an IP (e.g.: 172.16.100.10).
your_ports can be one port, multiple ports and even port ranges (e.g.: 554,8554,9000-9554)
your_manager can be either dumb or mysql but you probably want to use dumb. Check Cameradar's readme on the Docker Hub for more information.
tag allows you to specify a specific version for camerada. If you don't specify any tag, you will use the latest version by default (recommended)

The generated thumbnails will be in /tmp/thumbs on both your machine and the cameradar container.

For more complex use of the Docker image, see the Environment variables part of Cameradar's readme on the Docker Hub.

Quick install

The quick install uses docker to build Cameradar without polluting your machine with dependencies and makes it easy to deploy Cameradar in a few commands. However, it may require networking knowledge, as your docker containers will need access to the cameras subnetwork.

Dependencies

The only dependencies are docker, docker-tools, git and make.

Five steps guide

git clone https://github.com/EtixLabs/cameradar.git
Go into the Cameradar repository, then to the deployment directory
Tweak the conf/cameradar.conf.json as you need (see the onfiguration guide here for more information)
Run docker-compose build & docker-compose up

By default, the version of the package in the deployment should be the last stable release.

If you want to scan a different subnetwork or different ports, change the values CAMERAS_SUBNETWORKS and CAMERAS_PORTS in the docker-compose.yml file.

The generated thumbnails will be in the cameradar_thumbnails folder after cameradar has finished executing.

If you want to deploy your custom version of Cameradar using the same method, you should check the advanced docker deployment tutorial here.

Manual installation

The manual installation is recommended if you want to tweak Cameradar and quickly test them using CMake and running Cameradar in command-line. If you just want to use Cameradar, it is recommended to use the quick install instead.

Dependencies

To install Cameradar you will need these packages

cmake (cmake)
git (git)
gstreamer1.x (libgstreamer1.0-dev)
ffmpeg (ffmpeg)
boost (libboost-all-dev)
libcurl (libcurl4-openssl-dev)

Steps

The simplest way would be to follow these steps :

git clone https://github.com/EtixLabs/cameradar.git
mkdir build
cd build
cmake ..
make
cd cameradar_standalone
./cameradar -s the_subnet_you_want_to_scan

Advanced Docker deployment

In case you want to use Docker to deploy your custom version of Cameradar.

Dependencies

The only dependencies are docker and docker-compose.

Using the package generation script

git clone https://github.com/EtixLabs/cameradar.git
cd deployment
rm *.tar.gz
./build_last_package.sh
docker-compose build cameradar
docker-compose up cameradar

Deploy a custom version of Cameradar by hand

git clone https://github.com/EtixLabs/cameradar.git
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make package
cp cameradar_*_Release_Linux.tar.gz ../deployment
cd ../deployment
docker-compose build cameradar
docker-compose up cameradar

Configuration

Here is the basic content of the configuration file with simple placeholders :

{
  "mysql_db" : {
     "host" : "MYSQL_SERVER_IP_ADDRESS",
     "port" : MYSQL_SERVER_PORT,
     "user": "root",
     "password": "root",
     "db_name": "cmrdr"
  },
  "subnets" : "SUBNET1,SUBNET2,SUBNET3,[...]",
  "ports" : "PORT1,PORT2,[...]",
  "rtsp_url_file" : "conf/url.json",
  "rtsp_ids_file" : "conf/ids.json",
  "thumbnail_storage_path" : "/valid/path/to/a/storage/directory",
  "cache_manager_path" : "../cache_managers/dumb_cache_manager",
  "cache_manager_name" : "dumb"
}

This configuration is needed only if you want to overwrite the default values, which are :

{
  "subnets" : "localhost",
  "ports" : "554,8554",
  "rtsp_url_file" : "conf/url.json",
  "rtsp_ids_file" : "conf/ids.json",
  "thumbnail_storage_path" : "/tmp",
  "cache_manager_path" : "../cache_managers/dumb_cache_manager",
  "cache_manager_name" : "dumb"
}

This means that by default Cameradar will not use a database, will scan localhost and the ports 554 (default RTSP port) and 8554 (default emulated RTSP port), use the default constructor dictionaries and store the thumbnails in /tmp. If you need to override simply the subnets or ports, you can use the command line options.

The subnetworks should be passed separated by commas only, and their subnet format should be the same as used in nmap.

"subnets" : "172.100.16.0/24,172.100.17.0/24,localhost,192.168.1.13"

The RTSP ports for most cameras are 554, so you should probably specify 554 as one of the ports you scan. Not giving any ports in the configuration will scan every port of every host found on the subnetworks.

You can use your own files for the ids and routes dictionaries used to bruteforce the cameras, but the Cameradar repository already gives you a good base that works with most cameras.

The thumbnail storage path should be a valid and accessible directory in which the thumbnails will be stored.

The cache manager path and name variables are used to change the cache manager you want to load into Cameradar. If you want to, you can code your own cache manager using a database, a file, a remote server, [...]. Feel free to share it by creating a merge request on this repository if you developed a generic manager (It must not be specific to your company's infrastructure).

Output

For each camera, Cameradar will output these JSON objects :

{
   "address" : "173.16.100.45",
   "ids_found" : true,
   "password" : "123456",
   "path_found" : true,
   "port" : 554,
   "product" : "Vivotek FD9381-HTV",
   "protocol" : "tcp",
   "route" : "/live.sdp",
   "service_name" : "rtsp",
   "state" : "open",
   "thumbnail_path" : "/tmp/127.0.0.1/1463735257.jpg",
   "username" : "admin"
}

Check camera access

If you have VLC Media Player, you should be able to use the GUI to connect to the RTSP stream using this format : username:password@address:port/route

With the above result, the RTSP URL would be admin:123456@173.16.100.45:554/live.sdp

If you're still in your console however, you can go even faster by using vlc in commmand-line and just run vlc username:password@address:port/route with the camera's info instead of the placeholders.

Command line options

"-c" : Set a custom path to the configuration file (-c /path/to/conf)
"-s" : Set custom subnets (overrides configuration)
"-p" : Set custom ports (overrides configuration)
"-m" : Set number of threads (Default value : 1)
"-l" : Set log level
- "-l 1" : Log level DEBUG
  - Will print everything including debugging logs
- "-l 2" : Log level INFO
  - Prints every normal information
- "-l 4" : Log level WARNING
  - Only prints warning and errors
- "-l 5" : Log level ERROR
  - Only prints errors
- "-l 6" : Log level CRITICAL
  - Doesn't print anything since Cameradar can't have critical failures right now, however you can use this level to debug your own code easily or if you add new critical layers
"-d" : Launch the discovery tool
"-b" : Launch the bruteforce tool on all discovered devices
- Needs either to be launched with the -d option or to use an advanced cache manager (DB, file, ...) with data already present
"-t" : Generate thumbnails from detected cameras
- Needs either to be launched with the -d option or to use an advanced cache manager (DB, file, ...) with data already present
"-g" : Check if the stream can be opened with GStreamer
- Needs either to be launched with the -d option or to use an advanced cache manager (DB, file, ...) with data already present
"-v" : Display Cameradar's version
"-h" : Display this help
"--gst-rtsp-server" : Use this option if the bruteforce does not seem to work (only detects the username but not the path, or the opposite). This option will switch the order of the bruteforce to prioritize path over credentials, which is the way priority is handled for cameras that use GStreamer's RTSP server.

Under the hood

Cameradar uses nmap to map all of the subnetworks you specified in the configuration file (cameradar.conf.json), then parses its result to get all of the open RTSP streams that were detected.

After that, it uses cURL to send requests to the cameras and to try routes and ids for each camera until it is accessed or until all of the most used routes/ids (that you can modify in conf/ids.json and conf/url.json) were tried

Then, it uses FFMPEG to generate a lightweight thumbnail from the stream, which you could use to get a quick preview of the camera's view.

Finally, it tries to access the stream using a simple Gstreamer pipeline to check for the stream's encoding.

The output of Cameradar will be printed on the standard output and will also be accessible in the result.json file.

Cameradar uses nmap to map all of the subnetworks you specified in the configuration file (cameradar.conf.json), then parses its result to get all of the open RTSP streams that were detected.

Then, it uses FFMPEG to generate a lightweight thumbnail from the stream, which you could use to get a quick preview of the camera's view.

Finally, it tries to access the stream using a simple Gstreamer pipeline to check for the stream's encoding.

The output of Cameradar will be printed on the standard output and will also be accessible in the result.json file.

Contribution

Well there are many things we could code in order to add features to Cameradar. Adding other protocols than RTSP would be really cool, as well as making more generic cache managers. Improving Cameradar's performance or even the deployment could also be a great help!

If you're not into software development or not into C++, even updating the dictionaries would be a really cool contribution! Just make sure the ids and routes you add are default constructor credentials and not custom credentials.

If you have other cool ideas, feel free to share them with me at brendan.leglaunec@etixgroup.com !

Next improvements

Add a docker deployment to avoid the current deps hell
Development of a MySQL cache manager
Development of a JSON file cache manager
Development of an XML file cache manager
Make a standalone docker image
Push to DockerHub

Frequently Asked Questions

My camera's credentials are guessed by Cameradar but the RTSP url is not!

Your camera probably uses GST RTSP Server internally. Try the --gst-rtsp-server command-line option, and if it does not work, send me the cameradar output in DEBUG mode (-l 1) and I will help you.

Cameradar does not detect any camera!

That means that either your cameras are not streaming in RTSP or that they are not on the subnetwork you are scanning. In most cases, CCTV cameras will be on a private subnetwork. Use the -s option to specify your camera's subnetwork.

Cameradar detects my cameras, but does not manage to access them at all!

Maybe your cameras have been configured and the credentials / URL have been changed. Cameradar only guesses using default constructor values. However, you can use your own dictionary in which you just have to add your passwords. To do that, see how the configuration works.

It does not compile

You probably missed the part with the dependencies! Use the quick docker deployment, it will be easier and will not pollute your machine with useless dependencies! ;)

License

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

See the License for the specific language governing permissions and limitations under the License.