Path planning apparatus and path plan correction method of path planning apparatus

1 . A path planning apparatus for generating a path plan of a host vehicle from an image captured by a vehicle-mounted camera using a first deep learning model, recognizing left and right lane lines forming a travelling lane of the host vehicle from the image captured by the vehicle-mounted camera, and correcting the path plan using a recognition result of the lane lines, the path planning apparatus comprising:

a processor configured to:

in response to a distance in a lane width direction between a path of the vehicle in the path plan and a center position between the left and right lanes being equal to or greater than a predetermined distance, perform the correction so that the path of the vehicle approaches the center position;

determine whether the left and right lane lines satisfy a preset parallel condition;

not correct the path plan using the recognition result of the lane lines, in response to the left and right lane lines being determined as not satisfying the parallel condition, regardless of the distance between the path of the vehicle and the center position; and

control autonomous driving of the host vehicle in accordance with the path plan.

2 . The path planning apparatus of claim 1 , wherein the left and right lane lines are recognized using a second deep learning model from the image captured by the vehicle-mounted camera, and the second deep learning model is different from the first deep learning model.

3 . The path planning apparatus of claim 2 , wherein it is determined that the left and right lane lines do not satisfy the parallel condition when an angle formed by the left and right lane lines is greater than or equal to an allowable angle threshold, when the travelling lane is a straight line.

4 . The path planning apparatus of claim 2 , wherein a lane width of the travelling lane is calculated based on the recognition result of the lane lines whenever the recognition result of the lane lines is updated, a lane width difference that is a difference between the lane width before the update and the lane width after the update is calculated, and the left and right lane lines are determined as not satisfying the parallel condition when the lane width difference is greater than or equal to an allowable difference threshold.

5 . The path planning apparatus of claim 1 , wherein it is determined that the left and right lane lines do not satisfy the parallel condition when an angle formed by the left and right lane lines is greater than or equal to an allowable angle threshold, when the travelling lane is a straight line.

6 . The path planning apparatus of claim 1 , wherein a lane width of the travelling lane is calculated based on the recognition result of the lane lines whenever the recognition result of the lane lines is updated, a lane width difference that is a difference between the lane width before the update and the lane width after the update is calculated, and the left and right lane lines are determined as not satisfying the parallel condition when the lane width difference is greater than or equal to an allowable difference threshold.

7 . A path plan correction method of a path planning apparatus, the method comprising:

generating a path plan of a host vehicle from an image captured by a vehicle-mounted camera using a first deep learning model;

recognizing left and right lane lines forming a travelling lane of the host vehicle from the image captured by the vehicle-mounted camera;

correcting the path plan using a recognition result of the lane lines;

in response to a distance in a lane width direction between a path of the vehicle in the path plan and a center position between the left and right lanes being equal to or greater than a predetermined distance, performing the correction so that the path of the vehicle approaches the center position;

determining whether the left and right lane lines satisfy a preset parallel condition; and

not correcting the path plan using the recognition result of the lane lines, in response to the left and right lane lines being determined as not satisfying the parallel condition, regardless of the distance between the path of the vehicle and the center position; and

controlling autonomous driving of the host vehicle in accordance with the path plan.