US 12,283,120 B2
Method for detecting three-dimensional objects in relation to autonomous driving and electronic device
Chieh Lee, New Taipei (TW); Chih-Te Lu, New Taipei (TW); and Chin-Pin Kuo, New Taipei (TW)
Assigned to HON HAI PRECISION INDUSTRY CO., LTD., New Taipei (TW)
Filed by HON HAI PRECISION INDUSTRY CO., LTD., New Taipei (TW)
Filed on Aug. 25, 2022, as Appl. No. 17/895,496.
Claims priority of application No. 202210613590.X (CN), filed on May 31, 2022.
Prior Publication US 2023/0386231 A1, Nov. 30, 2023
Int. Cl. G06T 7/70 (2017.01); G06T 7/50 (2017.01); G06V 10/764 (2022.01); G06V 10/77 (2022.01); G06V 10/772 (2022.01); G06V 10/774 (2022.01); G06V 10/94 (2022.01); G06V 20/58 (2022.01); G06V 20/64 (2022.01)
CPC G06V 20/653 (2022.01) [G06T 7/50 (2017.01); G06T 7/70 (2017.01); G06V 10/764 (2022.01); G06V 10/7715 (2022.01); G06V 10/772 (2022.01); G06V 10/774 (2022.01); G06V 10/95 (2022.01); G06V 20/58 (2022.01); G06T 2207/10028 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30252 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for detecting three-dimensional (3D) objects in relation to autonomous driving comprising:
obtaining detection images and depth images corresponding to the detection images;
inputting the detection images into a trained object detection model, and determining object categories of objects in the detection images and two-dimensional (2D) bounding boxes of the objects by using the object detection model;
determining object models of the objects and 3D bounding boxes of the object models from a 3D object model library according to the object categories;
calculating point cloud data of the objects selected by the 2D bounding boxes according to the depth images and the 2D bounding boxes, and calculating distance from the depth camera to each of the object models;
determining rotation angles of the object models of the objects according to the object models of the objects and the point cloud data, comprising: obtaining first point cloud data of an outline of the objects according to the point cloud data; converting the object models of the objects into second point cloud data; generating a first plane by fitting contour points of the objects in the first point cloud data and calculating a curvature of the first plane; generating a second plane by fitting the points of the second point cloud data and calculating a curvature of the second plane; obtaining a curvature deviation value by calculating a difference between the curvature of the first plane and the curvature of the second plane, and determining the rotation angles of the object models of the objects according to the curvature deviation value;
determining positions of the objects in 3D space according to the distance from the depth camera to the object models, the rotation angles and the 3D bounding boxes.