Autonomous legorover... Using sensors and openCV for image analyze and obstacle detection. The openCV stream is sent to a mjpeg streamer that can be viewed in a webbrowser (firefox) on the same network.
Implementation: The implementation is using a simple game-development 'pattern' or strategy with a MainLoop. This loop around the different objects (Sensors, Vison, Driver) and all object have the methods initialize, update and draw. Each object, like the sensors may have more parts that are looped within this object. Sensors has a number of US distance sensors and a compass. The Driver has a base-class for 'state' with a couple of different implementations for different states. Vision is using classes for obstable-, face- and line-recognition wich data is tranfered to the driver-class.
LCM: Build: https://lcm-proj.github.io/build_instructions.html DataType: https://lcm-proj.github.io/tut_lcmgen.html Generate: lcm-gen -p usdistance.lcm