Run the program from the image-analyzer directory the filepaths are not absolute therefore a command should look like: ruby bin/imageanalyzer --task RUN
Usage: image_analyzer.rb [options]
--start N Search at the Nth result and start indexing at the Nth filename
--size N Search for and save the N images.
--task [DOWNLOAD|RUN] Download images or run images through pipeline.
--pipeline p1,p2,p3 Designate which parts of the pipeline to run. There is tesseract, laplace, sobel, and filter.
--query query Customize the search term to find new images. The default is credit cards.
ruby bin/image_analyzer.rb --task download --start 0 --size 20 --query 'drivers license'
ruby bin/image_analyzer.rb --task run --pipeline laplace,tesseract,filter
-
Tesseract needs to be able to be run from the command line. Follow their instructions to install.
-
OpenCV can be downloaded from http://opencv.org/downloads.html
-
It then needs to be compiled, instructions on Mac:
- Ensure CMake is installed ("sudo port install cmake")
- Create a build directory under the parent directory
- Run "cmake -G "Unix Makefiles" .."
- Run "make -j8"
- Run "sudo make install"
-
Install individual opencv processor
- rm -rf CMakeCache.txt CMakeFiles/ Makefile cmake_install.cmake sobel
- cmake .
- make
-
Lastly run
ruby install.rb
Run all unit tests.
rspec bin/*_spec.rb
- ruby gems
- google-search
- tesseract-ocr (interface for c program)
- Rmagick which requires port install imagemagick on Mac OSX
- c programs
- tesseract-ocr