python-test-server

A python test server to incorporate to the "CBIR visualization tool" and its server.

Prerequisits

Create a virtual enviroment how to create a virtual enviroment.

 virtualenv ~/BLCF
 source ~/BLCF/bin/activate

For python dependencies run:

 pip install --upgrade pip
 pip install -r requirements.txt

Make sure you are using theano order for the images. Edit '~/.keras/keras.json'

{    "image_dim_ordering": "th", "epsilon": 1e-07, "floatx": "float32", "backend": "theano"}

[NOTE] - In MAC_OS, matplotlib might cause problems using a python virtual enviroment. You might need to Create a file '~/.matplotlib/matplotlibrc' there and add the following code: 'backend: TkAgg'.

If you want the system to be evaluated, follow the next steps: 1. In the main project folder, create a file named compute_ap.cpp and copy/paste the code from HERE. 2. In your shell, from the same path, run the following code:
```
  ```
    > gcc compute_ap.cpp -o test -lstdc++
    > ./test
    > rm compute_ap.cpp
    > mv test compute_ap
  ```
```

Setting the paths and models

Set the path to the main folder in 'src/config.py'

MAIN_PATH = '/path/to/python-server/search_engine'

Download the data folder and place within the 'search_engine' folder.

search_engine/data

Execution

To run the code run:

python main.py

Documentation

Install apidoc

npm install -g apidoc

Run documentation generation script:

apidoc -o documentation -e node_modules

Open the file 'index.html' stored in the folder 'documentation' to see how to customize the system.

paulagd/python-test-server

python-test-server

Prerequisits

Setting the paths and models

Execution

Documentation