cvfeatures

image/video feature extraction toolkit, using OpenCV (cv2) depend on numpy

read input image/video or directory containing many images output json of various features

TODOs

• TODO 2012-12-18 decide on which set of features to extract (Lexing and Prithvi)
• TODO 2012-12-18 write an json feature reader to check the output (prithvi) this is in place of manually using a json validator http://jsonlint.com/
• TODO 2012-12-18 concatenating the json output has error right now
• TODO 2012-12-18 a number of command options does not work right now (zipped output, etc)

dir structure

test.sh -- test script, also see here for example use ./imgs -- a number of test imgs ./tmp -- where the test outputs go, its content should not be committed to git ./src -- source code dir, cvfeatures.py is the main script ./apps -- application / utility scripts for processing data, not part of the core (reusable) tools

misc

this seems a pretty good recipe for installing opencv on ubuntu 12.04 http://www.ozbotz.org/opencv-installation/