US 12,367,686 B2
Road structure detection method and apparatus
Wei Zhou, Shanghai (CN)
Assigned to Shenzhen Yinwang Intelligent Technologies Co., Ltd., Shenzhen (CN)
Filed by Shenzhen Yinwang Intelligent Technologies Co., Ltd., Shenzhen (CN)
Filed on Jun. 6, 2022, as Appl. No. 17/833,456.
Application 17/833,456 is a continuation of application No. PCT/CN2020/134225, filed on Dec. 7, 2020.
Claims priority of application No. 201911245257.2 (CN), filed on Dec. 6, 2019.
Prior Publication US 2022/0309806 A1, Sep. 29, 2022
Int. Cl. G06V 20/56 (2022.01); B60W 60/00 (2020.01)
CPC G06V 20/588 (2022.01) [B60W 60/001 (2020.02); B60W 2552/05 (2020.02); B60W 2552/10 (2020.02); B60W 2552/53 (2020.02)] 15 Claims
OG exemplary drawing
 
1. A road structure detection method comprising:
determining boundary information of a host lane, wherein the boundary information of the host lane represents a location of a boundary of the host lane;
determining boundary information of an adjacent lane based on information regarding trajectory of a vehicle and location of an edge of a road comprising the host and adjacent lanes,
wherein the boundary information of the adjacent lane represents a location of a boundary of the adjacent lane, and
wherein the boundary information of the adjacent lane comprises a location of one or more of (a) a left lane line (RL) of a right adjacent lane, (b) a right lane line (LR) of a left adjacent lane, (c) a left boundary of the right adjacent lane, and (d) a right boundary of the left adjacent lane,
wherein the location of the RL is obtained by (a) translating a road edge to the left by a first width that is an integer multiple of a lane width, (b) translating a vehicle trajectory to the left by a second width that is an odd multiple of a half-lane width, or (c) fusing the first location of the RL and the second location of the RL,
wherein the location of the LR is obtained by (a) translating the road edge to the right by a third width that is an integer multiple of the lane width, (b) translating the vehicle trajectory to the right by a fourth width that is an odd multiple of the half-lane width, or (c) fusing the fourth location of the LR and the fifth location of the LR,
wherein the location of the left boundary of the right adjacent lane is obtained by (a) translating the road edge to the left by a fifth width that is an integer multiple of the lane width, (b) translating a vehicle trajectory to the left by a sixth width that is an odd multiple of the half-lane width, or (c) fusing the seventh location of the left boundary and the eighth location of the left boundary using a preset algorithm, and
wherein the location of the right boundary of the left adjacent lane is obtained by (a) translating the road edge to the right by a seventh width that is an integer multiple of the lane width, (b) translating the vehicle trajectory to the right by an eighth width that is an odd multiple of the half-lane width, or (c) by fusing the tenth location of the right boundary and the eleventh location of the right boundary using the preset algorithm;
determining road structure information based on the boundary information of the host lane and the boundary information of the adjacent lane,
wherein the road structure information comprises location information of a merge point of the host lane and the adjacent lane, and/or a split point of the host lane and the adjacent lane,
generating a driving policy based on the road structure information; and
controlling driving of the vehicle based on action instructions associated with the driving policy.