This Repositories is contains code to run to WSP(Wound segementation prediction) web server. WSP is a wound analysis platform, focusing on pressure sore type wounds.
This docker compose will listen for:
- 80 (web service)
- 8080 (gui for database, phpMyAdmin.)
- 3306 (database service)
If your CPU not support tensorflow 2.6.0, Can using docker_image/wsp-server-compiler-tensorflow
to build self image.
- Web server php config at
web_server/config/php.ini
- SQL initialization script file at
db_server/init/
- about how to create SQL table, user.
In php.ini
attributes:
mysqli.default_host
,mysqli.default_user
,mysqli.default_pw
default DB config.ptmp.probability
per time useing the predict function, there is a certain chance to clear the temporary file.
if value is 100 means has 1/100 chance to clear.ptmp.path
predict temporary file store path.ptmp.tmpfile_lifetime
lifetime for temporary file, the unit is minutes.ptmp.tmpfile_max
maximum number of temporary file store.
The Predict mod pack is outside of this repository.
Download and then put in the following path.
web_server/WSP-Pages/wound
First install docker with docker-compose environment.
sudo apt-get install docker docker-compose
- Clone this repository.
- Download prediction module. module link
- Unzip the file, and then join prediction mod to
/web_server/WSP-Pages/wound
. - Using
chown -R 33:33 ./*
to change owner towww-data
. - Run
docker-compose up
- Web service is start.
Tip. You can build docker image in local, in docker_image directory, below is a build wsp-server image example:
$cd ./docker_images/wsp-server
$sudo docker build –t <imagename> .
Remenber change docker-compose file content, change image source.
currently directory will be like:
- mysql: store all database data.
- mysql_image: image storage path.
|---db_server---data---|---mysql
| |---mysql_image
|---images
|
|---wbe_server
|
|---docker-compose.yml