images_findpip_server.py is a python 3 script that can run as is, as a Docker container (docker run), or as a Docker service (docker service create).
Except for the special urls listed below, all output from this script must be redirected into a .tar.gz file. See the usage examples on how to redirect the output into a .tar.gz file.
The server listens on port 6003.
Currently, all files to be processed by this script must be on a web server or ftp server.
You will need to install opencv 3.4.0, python-numpy, python-scipy and a bunch of other packages. Use the opencv_install.sh file to install opencv 3.4.0 and python3.5.
The package manager used by opencv_install.sh is apt-get which limits the script to computers running the Ubuntu or Debian OS.
After opencv has been successfully installed, simply run python3.5 ./images_findpip.py.
Install Docker-ce on a Raspberry Pi with a single command by using their handy install script.
curl -sSL https://get.docker.com | sh
The above command should also work on most linux based computers. To install on Mac or Windows, visit https://docs.docker.com/install/.
Once Docker has been installed, you can then clone this repository, build the Docker image if desired, and run it as a container or service.
A pre-built image is availabe on hub.docker.com so that you do not need to build the image before using it.
To build the image, run docker build -t images_findpip . in the folder that you downloaded the contents of this repository to.
If you built the image use:
docker run -d -p 6003:6003 --network=host --name images_findpip images_findpip
To use the prebuilt image:
docker run -d -p 6003:6003 --network=host --name images_findpip mysmartbus/images_findpip
If you built the image use:
docker service create --publish published=6003,target=6003 --name images_findpip images_findpip
To use the prebuilt image:
docker service create --publish published=6003,target=6003 --name images_findpip mysmartbus/images_findpip
On a Raspberry Pi 2, using the prebuilt image took a little under 30 minutes to download and start.
There are three options that can be passed to the script to affect the data it returns. Note that the options reset to default values between calls to the script.
Multiple options can be specified in each request by separating them with three asterisks (***).
Example: debugfileon***returncolor
These options will be ignored if using one of the special urls listed below.
This option will turn off ALL debugging output. No data will be printed to the console or log file.
The script default is to turn debug mode on and send the output to the console.
You can view this output by running docker attach images_findpip if you started a container or by running docker logs images_findpip if you started a service.
Writes the debug log to a file named debug.txt which will be included in the .tar.gz file returned by the script.
The script default is not to include the debug.txt file.
If turned on, the pictures found by the script will be in color if the source image is in color.
Script default is to return grayscale (a.k.a black & white) pictures even if the source image is in color.
Has no effect if the source image is in grayscale.
Special URLs allow the client to retrieve a help/description file and the scripts version number. You do not need to redirect the output of these URLs unless you want to save the output to a text file.
http://helpme
Returns a text file containing instructions on how to use this script and the scripts version number.
http://version
Returns a text file containing only the scripts version number
These examples will work with the script regardless of how you started images_findpip.py.
You can try to send multiple urls per request but only the first url found will be processed. All other urls will be ignored.
Return color pictures.
curl -d "http://www.example.com/testimage01_color.png***returncolor~~~" <server_ip_or_hostname>:6003 > ~/results.tar.gz
The pictures found in testimage01_color.png will be in color with the script output saved to ~/results.tar.gz.
Return grayscale pictures.
curl -d "http://www.example.com/testimage01_color.png~~~" <server_ip_or_hostname>:6003 > ~/results.tar.gz
The pictures found in testimage01_color.png will be converted to grayscale with the script output saved to ~/results.tar.gz.
Return grayscale pictures and the debug.txt file.
curl -d "http://www.example.com/testimage01_color.png***debugfileon~~~" <server_ip_or_hostname>:6003 > ~/results.tar.gz
The pictures found in testimage01_color.png will be converted to grayscale with the script output saved to ~/results.tar.gz. Also included in results.tar.gz will be a file named debug.txt showing what functions the script ran to get the pictures that it found.
This is a valid request.
It demonstrates what happens when you include multiple urls, random characters and multiple options. Also note that some options have extra (more than 3) asterisks before and/or after them.
curl -d "abcd***http://www.example.com/some/folder/testimage01_color.png*****returncolor***http://www.example.com/testimage45.png****debugfileon***returncolor~~~" <server_ip_or_hostname>:6003 > ~/some/folder/somwhere/testimage01_color.tar.gz
The pictures found in testimage01_color.png will be in color. Also included in results.tar.gz will be a file named debug.txt showing what functions the script ran to get the pictures that it found. The scripts output will be saved to ~/some/folder/somwhere/testimage01_color.tar.gz
The url http://www.example.com/testimage45.png will be ignored.
See the files in the examples folder.