cvisionai / oceancv Goto Github PK

OceanCV provides computer vision algorithms and tools for underwater image analysis. This includes image processing, pattern recognition, machine learning and geometric algorithms but also functionality for navigation data processing, data provenance etc.

OceanCV bundles marine image analysis functionality. Most of the algorithms are related to publications in scientific journals. See www.oceancv.com for details.

OceanCV uses cmake. So follow the usual: mkdir build && cd build && cmake ..

You can enable / disable the following build targets:

• WITH_MPEG7 : Build MPEG7 feature extraction library depencendy
• WITH_CUDA : Build with OpenCV GPU functionality
• WITH_TESTS : Build tests (uses Google Test)
• WITH_EXAMPLES : Build examples
• WITH_BIN : Build binary tools
• WITH_DOC : Build doxygen documentation

By default, all targets are being built. And as always, default install path is /usr/local, but you can control this using cmake -DCMAKE_INSTALL_PREFIX=/target/path ..

Dr. Timm Schoening

tschoening [at] geomar [dot] de

DeepSea Monitoring Group

GEOMAR Helmholtz-Centre for Ocean Research Kiel

• OpenCV (required, ideally with CUDA)
• jsoncpp (required)
• curl (required)
• MPEG7FexLib (optional) -> included and provided by Muhammet Bastan (Bilkent University) - many thanks!
• GTest (optional, for tests)
• doxygen (optional, for documentation)