Lane detection and distance estimation using single-view geometry
Disclosed are methods, devices, and computer-readable media for detecting lanes and objects in image frames of a monocular camera. In one embodiment, a method is disclosed comprising receiving a sample set of image frames; detecting a plurality of markers in the sample set of image frames using a convolutional neural network (CNN); fitting lines based on the plurality of markers; detecting a plurality of vanishing points based on the lines; identifying a best fitting horizon for the sample set of image frames via a RANSAC algorithm; computing an inverse perspective mapping (IPM) based on the best fitting horizon; and computing a lane width estimate based on the sample set of image frames using the IPM in a rectified view and the parallel line fitting.
1. A method comprising:
detecting vanishing points in image frames based on lines fitted using lane markers detected in the image frames;
identifying a best-fitting horizon line for the image frames using the vanishing points;
computing an inverse perspective mapping of the image frames using the best-fitting horizon line; and
computing a lane width estimate using the inverse perspective mapping and the lines fitted using lane markers detected in the image frames.
2. The method of claim 1 , wherein computing the inverse perspective mapping comprises computing a rectification homography based on an intrinsic matrix of a camera and a rotation matrix calculated based on the best-fitting horizon line and a road plane normal.
3. The method of claim 1 , further comprising computing a height of a camera that capture the image frames based on a lane width.
4. The method of claim 1 , further comprising rectifying the image frames using the inverse perspective mapping to generate a rectified view.
5. The method of claim 4 , further comprising predicting at least one missing lane boundary using the rectified view.
6. The method of claim 1 , further comprising:
receiving a second set of image frames;
rectifying markers detected in the second set of image frames using the inverse perspective mapping; and
fitting a plurality of lines to the markers.
7. The method of claim 1 , further comprising:
detecting an object in the image frames; and
calculating a distance from a camera to the object based on a camera height and a normal plane.
8. A non-transitory computer-readable storage medium for tangibly storing computer program instructions capable of being executed by a computer processor, the computer program instructions defining steps of:
detecting vanishing points in image frames based on lines fitted using lane markers detected in the image frames;
identifying a best-fitting horizon line for the image frames using the vanishing points;
computing an inverse perspective mapping of the image frames using the best-fitting horizon line; and
computing a lane width estimate using the inverse perspective mapping and the lines fitted using lane markers detected in the image frames.
9. The non-transitory computer-readable storage medium of claim 8 , wherein computing the inverse perspective mapping comprises computing a rectification homography based on an intrinsic matrix of a camera and a rotation matrix calculated based on the best-fitting horizon line and a road plane normal.
10. The non-transitory computer-readable storage medium of claim 8 , further comprising computing a height of a camera that capture the image frames based on a lane width.
11. The non-transitory computer-readable storage medium of claim 8 , further comprising rectifying the image frames using the inverse perspective mapping to generate a rectified view.
12. The non-transitory computer-readable storage medium of claim 11 , further comprising predicting at least one missing lane boundary using the rectified view.
13. The non-transitory computer-readable storage medium of claim 8 , the steps further comprising:
receiving a second set of image frames;
rectifying markers detected in the second set of image frames using the inverse perspective mapping; and
fitting a plurality of lines to the markers.
14. The non-transitory computer-readable storage medium of claim 8 , the steps further comprising:
detecting an object in the image frames; and
calculating a distance from a camera to the object based on a camera height and a normal plane.
15. A device comprising:
a processor; and
a storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising:
logic, executed by the processor, for detecting vanishing points in image frames based on lines fitted using lane markers detected in the image frames,
logic, executed by the processor, for identifying a best-fitting horizon line for the image frames using the vanishing points,
logic, executed by the processor, for computing an inverse perspective mapping of the image frames using the best-fitting horizon line, and
logic, executed by the processor, for computing a lane width estimate using the inverse perspective mapping and the lines fitted using lane markers detected in the image frames.
16. The device of claim 15 , wherein computing the inverse perspective mapping comprises computing a rectification homography based on an intrinsic matrix of a camera and a rotation matrix calculated based on the best-fitting horizon line and a road plane normal.
17. The device of claim 15 , further comprising computing a height of a camera that capture the image frames based on a lane width.
18. The device of claim 15 , further comprising rectifying the image frames using the inverse perspective mapping to generate a rectified view.
19. The device of claim 18 , further comprising predicting at least one missing lane boundary using the rectified view.
20. The device of claim 15 , the program logic further comprising:
logic, executed by the processor, for receiving a second set of image frames;
logic, executed by the processor, for rectifying markers detected in the second set of image frames using the inverse perspective mapping; and
logic, executed by the processor, for fitting a plurality of lines to the markers.