Method and apparatus for modeling object, storage medium, and vehicle control method
Disclosed are a method and apparatus for modeling an object, a storage medium, and a vehicle control method. The method for modeling an object includes obtaining a point cloud associated with surroundings acquired by a perception system so that points corresponding to each part of the object are respectively recognized from the point cloud, and then corresponding representations are respectively generated for the object wholly and each part thereof. As such, the object is modeled and the modeling precision is improved, whereby a movement state of the object may be accurately determined.
1 . A method, comprising:
obtaining a point cloud associated with surroundings of a perception system;
determining, from the point cloud, a group of points associated with an object and a first subset and a second subset of the group of points;
determining, based on the first subset and the second subset, a first confidence that the object comprises a first part associated with the first subset and a second confidence that the object comprises a second part associated with the second subset;
determining whether at least two parts of an articulated object which are articulated with each other are sensed by the perception system based on the first confidence and the second confidence; and
in response to both the first confidence and the second confidence being above a predetermined threshold, determining that the at least two parts of the articulated object are sensed by the perception system and generating a hierarchical representation of the object, comprising a first representation of the object, a second representation of the first part, and a third representation of the second part, the second and third representations being at least partially located within the first representation.
2 . The method according to claim 1 , further comprising:
determining a third subset and a fourth subset comprised in the second subset of the group of points;
determining, based on the third subset and the fourth subset, a third confidence that the object comprises a third part associated with the third subset and a fourth confidence that the object comprises a fourth part associated with the fourth subset;
generating a fourth representation of the third part and a fifth representation of the fourth part in response to both the third and fourth confidences being above the predetermined threshold, the fourth and fifth representations being at least partially located within the third representation.
3 . The method according to claim 2 , wherein
the first representation has a first geometry, the second representation has a second geometry, the third representation has a third geometry, the fourth representation has a fourth geometry, and the fifth representation has a fifth geometry.
4 . The method according to claim 3 , wherein
at least one of the first geometry, the second geometry, the third geometry, the fourth geometry, and the fifth geometry is rectangular.
5 . The method according to claim 1 , wherein determining, from the point cloud, the group of points associated with the object and the first subset and the second subset of the group of points comprises:
determining a type of the object based on the group of points; and
determining the first subset and the second subset of the group of points in response to the type of the object being a predetermined type.
6 . The method according to claim 1 , wherein determining, from the point cloud, the group of points associated with the object and the first subset and the second subset of the group of points comprises:
performing feature detection on the point cloud to determine the group of points associated with the object from the point cloud; and
performing feature detection on the group of points to determine the first subset of the group of points that is associated with the first part and the second subset of the group of points that is associated with the second part.
7 . The method according to claim 1 , wherein determining, from the point cloud, the group of points associated with the object and the first subset and the second subset of the group of points comprises:
determining a part of the point cloud based on image data associated with the object; and
performing feature detection on the part of the point cloud to determine, from the point cloud, the group of points associated with the object, the first subset of the group of points that is associated with the first part, and the second subset of the group of points that is associated with the second part.
8 . The method according to claim 1 , wherein the second and third representations have a first positional relationship in response to the object being in a first state and have a second positional relationship in response to the object being in a second state, and the first positional relationship is different from the second positional relationship.
9 . The method according to claim 8 , wherein the first representation has a first size and the second and third representations are located within the first representation in response to the object being in the first state, and
wherein the first representation has a second size and the second and third representations are located within the first representation in response to the object being in the second state.
10 . The method according to claim 1 , wherein the object comprises a truck, the first part comprises a tractor, and the second part comprises a trailer.
11 . The method according to claim 1 , further comprising:
generating a trajectory for the vehicle to travel along at least partly based on the hierarchical representation of the object; and
controlling the vehicle to travel along the generated trajectory.
12 . A method, comprising:
obtaining a point cloud associated with surroundings of a perception system;
determining, from the point cloud, a first group of points that are associated with a first object in the surroundings and a second group of points that are associated with a second object in the surroundings;
determining, based on the first group of points and the second group of points, whether the first object is within a predetermined distance from the second object to determine whether the first object is articulated to the second object;
generating a first representation of the first object and the second object, a second representation of the first object, and a third representation of the second object in response to the first object being articulated to the second object, the second representation and the third representation being at least partially located within the first representation.
13 . The method according to claim 12 , further comprising:
determining a velocity relationship between the first object and the second object based on the first group of points and the second group of points; and
determining that the first object is articulated to the second object in response to the first object being within the predetermined distance from the second object and a difference in velocity between the first object and the second object being less than a predetermined threshold.
14 . The method according to claim 12 , further comprising:
generating a trajectory for the vehicle to travel along at least partly based on the hierarchical representation of the object; and
controlling the vehicle to travel along the generated trajectory.
15 . An electronic device, comprising: a processor and a memory, wherein the memory stores program instructions executable by the processor to implement a method comprising:
obtaining a point cloud associated with surroundings of a perception system;
determining, from the point cloud, a group of points associated with an object and a first subset and a second subset of the group of points;
determining, based on the first subset and the second subset, a first confidence that the object comprises a first part associated with the first subset and a second confidence that the object comprises a second part associated with the second subset;
determining whether at least two parts of an articulated object which are articulated with each other are sensed by the perception system based on the first confidence and the second confidence; and
in response to both the first confidence and the second confidence being above a predetermined threshold, determining that the at least two parts of the articulated object are sensed by the perception system and generating a hierarchical representation of the object, comprising a first representation of the object, a second representation of the first part, and a third representation of the second part, the second and third representations being at least partially located within the first representation.
16 . The electronic device according to claim 15 , wherein determining, from the point cloud, the group of points associated with the object and the first subset and the second subset of the group of points comprises:
performing feature detection on the point cloud to determine the group of points associated with the object from the point cloud; and
performing feature detection on the group of points to determine the first subset of the group of points that is associated with the first part and the second subset of the group of points that is associated with the second part.
17 . The electronic device according to claim 15 , wherein determining, from the point cloud, the group of points associated with the object and the first subset and the second subset of the group of points comprises:
determining a part of the point cloud based on image data associated with the object; and
performing feature detection on the part of the point cloud to determine, from the point cloud, the group of points associated with the object, the first subset of the group of points that is associated with the first part, and the second subset of the group of points that is associated with the second part.
18 . The electronic device according to claim 15 , wherein the second and third representations have a first positional relationship in response to the object being in a first state and have a second positional relationship in response to the object being in a second state, and the first positional relationship is different from the second positional relationship.
19 . The electronic device according to claim 18 , wherein the first representation has a first size and the second and third representations are located within the first representation in response to the object being in the first state, and
wherein the first representation has a second size and the second and third representations are located within the first representation in response to the object being in the second state.
20 . The electronic device according to claim 15 , wherein the object comprises a truck, the first part comprises a tractor, and the second part comprises a trailer.