Image processing device
Provided is an image processing device that, by performing highly accurate correction of a parallax error caused by a vertical offset between left and right cameras, makes it possible to accurately distinguish between a control object such as an object dropped on a road surface and road surface paint, and accurately detect a control object having a low step. The image processing device includes a specific region extraction unit that extracts a specific region on a road surface from a plurality of images captured by each of a plurality of in-vehicle cameras; a road surface parameter estimation unit that estimates a road surface structure represented by a road surface parameter, from a road surface portion other than the specific region, the road surface portion being a road surface in the image; a specific region parallax estimation unit that obtains, from the road surface structure, an ideal parallax to be obtained in the specific region; a parallax image generation unit that obtains a parallax from the plurality of images; and a vertical offset calculation unit that obtains a vertical offset amount of the plurality of images by using an actual parallax of the specific region, the ideal parallax of the specific region, and arrangement information in the image of the specific region.
1 . An image processing device, comprising:
a specific region extraction unit that extracts a specific region on a road surface from a plurality of images captured by each of a plurality of in-vehicle cameras;
a road surface parameter estimation unit that estimates a road surface structure represented by a road surface parameter, from a road surface portion other than the specific region, the road surface portion being a road surface in the image;
a specific region parallax estimation unit that obtains, from the road surface structure, an ideal parallax to be obtained in the specific region;
a parallax image generation unit that obtains a parallax from the plurality of images; and
a vertical offset calculation unit that obtains a vertical offset amount of the plurality of images by using an actual parallax of the specific region, the ideal parallax of the specific region, and arrangement information in the image of the specific region,
wherein the specific region is a white line region on a road surface, and
wherein the vertical offset calculation unit calculates the vertical offset amount from a difference between a parallax of the white line region and a parallax of the road surface portion and an inclination angle of a white line in the image.
2 . An image processing device, comprising:
a specific region extraction unit that extracts a specific region on a road surface from a plurality of images captured by each of a plurality of in-vehicle cameras;
a road surface parameter estimation unit that estimates a road surface structure represented by a road surface parameter, from a road surface portion other than the specific region, the road surface portion being a road surface in the image;
a specific region parallax estimation unit that obtains, from the road surface structure, an ideal parallax to be obtained in the specific region;
a parallax image generation unit that obtains a parallax from the plurality of images; and
a vertical offset calculation unit that obtains a vertical offset amount of the plurality of images by using an actual parallax of the specific region, the ideal parallax of the specific region, and arrangement information in the image of the specific region,
wherein the parallax image generation unit corrects optical distortion of an image captured by the in-vehicle camera, and
wherein a vertical offset of a plurality of images captured by the in-vehicle camera is corrected on the basis of the vertical offset amount.
3 . The image processing device according to claim 2 ,
wherein the parallax image generation unit determines the feasibility of executing the correction and the correction amount, from chronological information on the vertical offset amount estimated in the past.
4 . The image processing device according to claim 2 ,
wherein the specific region extraction unit extracts an oblique line region on the road surface.
5 . The image processing device according to claim 2 ,
wherein the specific region extraction unit extracts at least one linear region among a white line, road surface paint, and a shadow on the road surface.
6 . The image processing device according to claim 3 ,
wherein the road surface parameter estimation unit estimates the road surface parameter from an image captured by the in-vehicle camera, and
wherein the specific region parallax estimation unit estimates the ideal parallax of the specific region on the basis of the road surface parameter.
7 . The image processing device according to claim 5 ,
wherein the arrangement information is an inclination angle of the specific region in the image.
8 . The image processing device according to claim 7 ,
wherein the vertical offset amount calculation unit determines whether to calculate the vertical offset amount on the basis of the reliability of the road surface parameter.
9 . An image processing method, comprising the steps of:
extracting a specific region on a road surface from a plurality of images captured by each of a plurality of in-vehicle cameras;
estimating a road surface structure represented by a road surface parameter, from a road surface portion other than the specific region, the road surface portion being a road surface in the image;
obtaining, from the road surface structure, an ideal parallax to be obtained in the specific region;
obtaining a parallax from the plurality of images; and
obtaining a vertical offset amount of the plurality of images by using an actual parallax of the specific region, the ideal parallax of the specific region, and arrangement information in the image of the specific region,
wherein the specific region is a white line region on a road surface, and
wherein obtaining the vertical offset amount of the plurality of images further comprises calculating the vertical offset amount from a difference between a parallax of the white line region and a parallax of the road surface portion and an inclination angle of a white line in the image.