Vehicle vision system including camera to ground alignment using bird's eye view image
A vision system for a vehicle includes a camera configured to generate one of distorted and undistorted images. An alignment system is configured to generate a region of interest (ROI) for the camera, generate a bird's eye view (BEV) homography to convert undistorted points in the ROI to BEV points, generate a BEV image based on the one of distorted and undistorted images and one of the BEV homography and an updated BEV homography, adaptively update the BEV homography to generate the updated BEV homography, detect current locations of features in the BEV image, predict future locations of the features for a next BEV image, compare the current locations of the features in the BEV image with prior predicted locations, and convert the BEV points in the BEV image to an undistorted points in a perspective undistorted image using an inverse of the updated BEV homography.
1. A vision system for a vehicle comprising:
a camera configured to generate one of distorted and undistorted images; and
an alignment system configured to:
generate a region of interest (ROI) for the camera;
generate a bird's eye view (BEV) homography to convert undistorted points in the ROI to BEV points;
generate a BEV image based on the one of distorted and undistorted images and one of the BEV homography and an updated BEV homography;
adaptively update the BEV homography to generate the updated BEV homography;
detect current locations of features in the BEV image;
predict future locations of the features for a next BEV image;
compare the current locations of the features in the BEV image with prior predicted locations; and
convert the BEV points in the BEV image to an undistorted points in a perspective undistorted image using an inverse of the updated BEV homography.
2. The vision system of claim 1 , wherein the alignment system is configured to adaptively update the BEV homography by calculating alignment of the camera based on the detected current locations and the prior predicted locations.
3. The vision system of claim 2 , wherein the alignment system is configured to adaptively update the BEV homography by determining a plurality of feature alignment values based on the detected current locations and the prior predicted locations.
4. The vision system of claim 3 , wherein the alignment system is configured to adaptively update the BEV homography by calculating a deviation measurement of the plurality of feature alignment values.
5. The vision system of claim 4 , wherein the alignment system is configured to adaptively update the BEV homography by calculating a function of the plurality of feature alignment values when the deviation measurement is less than a predetermined threshold.
6. The vision system of claim 5 , wherein:
the alignment system is configured to use the function to update one of the BEV homography and the updated BEV homography when the deviation measurement is less than a predetermined threshold, and
not update one of the BEV homography and the updated BEV homography when the deviation measurement is greater than a predetermined threshold.
7. The vision system of claim 2 , wherein the alignment system is configured to adaptively update a feature tracking condition prior to calculating alignment of the camera based on the detected current locations and the prior predicted locations.
8. The vision system of claim 2 , wherein the alignment system is configured to adaptively update an enabling condition prior to calculating alignment of the camera based on the detected current locations and the prior predicted locations.
9. The vision system of claim 1 , wherein the alignment system is configured to use a first ROI during one portion of an alignment iteration and a second ROI during another portion of an alignment iteration, wherein the first ROI is different than the second ROI.
10. The vision system of claim 1 , wherein the alignment system is configured to use an adaptive variable based on a default variable, a relaxed variable, a number of results generated, and a predetermined number of results for adaptation.
11. A method for aligning a camera of a vision system for a vehicle to ground, comprising:
generating a region of interest (ROI) for a camera;
generating a bird's eye view (BEV) homography to convert undistorted points in the ROI to BEV points;
generating a BEV image based on the one of distorted and undistorted images and one of the BEV homography and an updated BEV homography;
adaptively updating the BEV homography to generate the updated BEV homography;
detecting current locations of features in the BEV image;
predicting future locations of the features for a next BEV image;
comparing the current locations of the features in the BEV image with prior predicted locations; and
converting the BEV points in the BEV image to an undistorted points in a perspective undistorted image using an inverse of the updated BEV homography.
12. The method of claim 11 , further comprising adaptively updating the BEV homography by calculating alignment of the camera based on the detected current locations and the prior predicted locations.
13. The method of claim 12 , further comprising adaptively updating the BEV homography by determining a plurality of feature alignment values based on the detected current locations and the prior predicted locations.
14. The method of claim 13 , further comprising adaptively updating the BEV homography by calculating a deviation measurement of the plurality of feature alignment values.
15. The method of claim 14 , further comprising adaptively updating the BEV homography by calculating a function of the plurality of feature alignment values when the deviation measurement is less than a predetermined threshold.
16. The method of claim 15 , further comprising:
using the function to update one of the BEV homography and the updated BEV homography when the deviation measurement is less than a predetermined threshold, and
not updating one of the BEV homography and the updated BEV homography when the deviation measurement is greater than a predetermined threshold.
17. The method of claim 12 , further comprising adaptively updating a feature tracking condition prior to calculating alignment of the camera based on the detected current locations and the prior predicted locations.
18. The method of claim 12 , further comprising adaptively updating an enabling condition prior to calculating alignment of the camera based on the detected current locations and the prior predicted locations.
19. The method of claim 11 , further comprising using a first ROI during one portion of an alignment iteration and a second ROI during another portion of an alignment iteration, wherein the first ROI is different than the second ROI.
20. The method of claim 11 , further comprising using an adaptive variable based on a default variable, a relaxed variable, a number of results generated, and a predetermined number of results for adaptation.