Real-time robust surround view parking space detection and tracking

1. A method, comprising:

receiving a plurality of camera images from a plurality of cameras attached to a vehicle in motion;

stitching the plurality of camera images to simulate a surround view of the vehicle;

recognizing at last one potential parking space within the surround view of the vehicle;

determining an available free-space via the surround view of the vehicle to indicate which area is safest to drive to the at last one potential parking space;

determining a percentage of free-space within the at last one potential parking space to indicate whether the at last one potential parking space is available;

estimating a motion parameter by camera motion estimation of the plurality of cameras;

tracking the at last one potential parking space based on the motion parameter to indicate a relative position of the vehicle and the at last one potential parking space as tracklets;

determining a location of the at least one potential parking space;

determining an amount of free space surrounding the at last one potential parking space;

determining a direction from a vehicle location to the at last one potential parking space;

determining a distance from the vehicle location to the at last one potential parking space;

determining two entrance corners of the at least one potential parking space; and

clustering the at least one potential parking space into a set of clusters indicating entrance points.

2. The method of claim 1 , further comprising updating the location of the at last one potential parking space based on the camera motion estimation.

3. The method of claim 1 , further comprising determining a stopping point from the vehicle location to the at last one potential parking space.

4. The method of claim 1 , further comprising segmenting the set of clusters of the at last one potential parking space.

5. The method of claim 4 , further comprising ranking the set of clusters based on combinatorial optimization.

6. The method of claim 5 , further comprising selecting one of the set of clusters based on distance thresholding.

7. A non-transitory computer readable medium comprising instructions that, when read by a processor, cause the processor to perform:

receiving a plurality of camera images from a plurality of cameras attached to a vehicle in motion;

stitching the plurality of camera images to simulate a surround view of the vehicle;

recognizing at last one potential parking space within the surround view of the vehicle;

determining an available free-space via the surround view of the vehicle to indicate which area is safest to drive to the at last one potential parking space;

determining a percentage of free-space within the at last one potential parking space to indicate whether the at last one potential parking space is available;

estimating a motion parameter by camera motion estimation of the plurality of cameras;

tracking the at last one potential parking space based on the motion parameter to indicate a relative position of the vehicle and the at last one potential parking space as tracklets;

determining a location of the at last one potential parking space;

determining an amount of free space surrounding the at last one potential parking space;

determining a direction from a vehicle location to the at last one potential parking space;

determining a distance from the vehicle location to the at last one potential parking space;

determining two entrance corners of the at last one potential parking space;

clustering the at last one potential parking space into a set of clusters indicating entrance points; and

selecting one of the set of clusters based on distance thresholding.

8. The non-transitory computer readable medium of claim 7 , further comprising segmenting the set of clusters of the at last one potential parking space.

9. The non-transitory computer readable medium of claim 8 , further comprising ranking the set of clusters based on combinatorial optimization.

10. The non-transitory computer readable medium of claim 9 , wherein the ranking is based on a Hungarian algorithm.

11. The non-transitory computer readable medium of claim 7 , wherein the surround view of the vehicle is a bird's eye view.