OSPRAI
Open Source Picture Review & Analysis Investigator
A work in progress modular open source media categoriser and analysis tool, developed in Python 2.7, GTK 3, and Glade 3.18.
Development Enviroment:
Windows
For those who wish to develop this project on Microsoft Windows, you will require the All-In-One PyGI/PyGObject Installer which can be obtained here
The "dir_parser" module is dependant on: OpenCV 2.x library - Available here Example: pip install opencv_python‑2.4.13.2‑cp27‑cp27m‑win32.whl
NumPy Example: pip install numpy
XXHash Library - Available here
Ubuntu
You can install OpenCV from the Ubuntu or Debian repository via the command: sudo apt-get install libopencv-dev python-opencv
MacOS
To develop on MacOS you must first install the Homebrew package manager for MacOS which is obtained by running the following command: /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
You will then need to use homebrew and pip to install the following packages (tested using MacOS 10.12 - Sierra), the commands are listed as follows (the system may need to be restarted before running Osprai):
- brew install gtk+3
- brew install pygobject3
- brew tap homebrew/science
- brew install opencv
- pip install numpy
- brew install xxhash
- pip install xxhash
Sample Data:
It is accepted that developers may require large sample datasets for development. Refer to "sample_data.py" for a simple script that crawls NASA's Astronomy Picture of the Day archive and downloads the images. The author of this code is Cecil Woebker