Repository for UCSD AUVSI computer vision software suite

cmake libopencv-dev libzmq-dev python-dev libboost-python-dev libtesseract-dev libleptonica-dev

Optional: QT widgets 5

• GPSLatitude
• GPSLongitude
• GPSAltitude (relative to Mean Sea Level MSL)
• GPSSpeed hack: should contain mean ground level altitude (assuming a flat field)
• GPSImgDirection compass direction in degrees where 0 == North, 90 == East, 180 == South, 270 == West

see Backbone/IMGData.hpp

the actual images are encoded to vectors of unsigned chars (i.e. uint8_t) so need to be decoded back to cv::Mat if you want to process them

see Backbone/DistMaps.cpp

note that you will not have to modify this unless you wrote C++ code or want to reorganize the flowchart

Heimdall Vision Flowchart 2015:

• Salient Object Detection
• QR Code Read/Check
• Segmentation
• OCR
• Results Aggregation

Not supported: Python-Shape-Recognition (though the current C++ implementation is good)