Optical state estimation and simulation environment for unmanned aerial vehicles

A system comprising:

• a first processor configured to:
• simulate an unmanned aerial vehicle (UAV),
• estimate pitch, roll and groundspeed associated with said UAV, based at least in part, on a first simulated view from a graphical simulator configured to simulate at least one visual sensor, wherein said pitch and said roll are based on an angle of said simulated UAV relative to a simulated horizon, said simulated horizon determined by a Hough transformation of an output of a Sobel edge detection filter applied to said first simulated view, and wherein said groundspeed is based on a block matching algorithm configured to match a region of said first simulated view to a region of a previous simulated view and to estimate motion based on a difference in position of said region of said first simulated view and said region of said previous simulated view, wherein a size of said region is based on a maximum groundspeed, altitude and field of view of said simulated visual sensor,
• compare said estimated pitch, roll and groundspeed, to data generated by a simulated global positioning system (GPS), said comparison to evaluate said simulated visual sensor, and
• determine at least one of a position and an orientation of said UAV, based at least in part, on said estimated pitch, roll and groundspeed;
• and
• a second processor configured to:
• simulate an environment,
• simulate said UAV in said environment at said at least one of said position and said orientation;
• generate a second simulated view based on at least one of said position and said orientation;
• and
• provide said second simulated view to said first processor to repeat said estimation of said pitch, roll and groundspeed using said second simulated view.

An article comprising a non-transitory storage medium having stored thereon instructions that when executed by a machine result in the following operations for simulating an unmanned aerial vehicle (UAV) in a simulated environment:

• (1) receiving a first simulated view of said simulated environment from a graphical simulator configured to simulate at least one visual sensor positioned on said simulated UAV;
• (2) estimating pitch, roll and groundspeed associated with said simulated UAV, based at least in part, on said first simulated view from said simulated visual sensor, using a first processor, wherein said pitch and said roll are based on an angle of said simulated UAV relative to a simulated horizon in said simulated environment, said simulated horizon determined by a Hough transformation of an output of a Sobel edge detection filter applied to said first simulated view, and wherein said groundspeed is based on a block matching algorithm configured to match a region of said first simulated view to a region of a previous simulated view and to estimate motion based on a difference in position of said region of said first simulated view and said region of said previous simulated view, wherein a size of said region is based on a maximum groundspeed, altitude and field of view of said simulated visual sensor;
• (3) comparing said estimated pitch, roll and groundspeed, to data generated by a simulated global positioning system (GPS), said comparison to evaluate said simulated visual sensor;
• (4) determining at least one of a position and an orientation of said simulated UAV, based at least in part, on said estimated pitch, roll and groundspeed, using said first processor;
• (5) simulating said UAV in said simulated environment at said at least one of said position and said orientation, using a second processor;
• (6) generating a second simulated view based on at least one of said position and said orientation;
• and
• (7) repeating steps 1 through 5 using said second simulated view.