US 12,092,470 B2
Vehicle localization method and device, electronic device and storage medium
Chao Wang, Beijing (CN); Chi Zhang, Beijing (CN); Hao Liu, Beijing (CN); Zhijun Xie, Beijing (CN); Kuiyuan Yang, Beijing (CN); Kun Guo, Beijing (CN); Rui Cai, Beijing (CN); and Zhiwei Li, Beijing (CN)
Assigned to BEIJING XIAOMI MOBILE SOFTWARE CO., LTD., Beijing (CN)
Filed by BEIJING XIAOMI MOBILE SOFTWARE CO., LTD., Beijing (CN)
Filed on Jan. 28, 2022, as Appl. No. 17/588,076.
Claims priority of application No. 202111089405.3 (CN), filed on Sep. 16, 2021.
Prior Publication US 2023/0078721 A1, Mar. 16, 2023
Int. Cl. G01C 21/34 (2006.01); G06V 20/56 (2022.01); G06V 20/58 (2022.01)
CPC G01C 21/3476 (2013.01) [G01C 21/3461 (2013.01); G06V 20/58 (2022.01); G06V 20/588 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A vehicle localization method, comprising:
acquiring an image of a scene where a vehicle is located, the image comprising a semantic element;
determining an element category corresponding to the semantic element, wherein the element category comprises at least one member selected from the group consisting of a discrete landmark, a continuous thin landmark and a continuous thick landmark;
matching the semantic element with a map element to acquire a matching result by a matching method corresponding to the element category; and
determining localization information of the vehicle according to the matching result;
wherein if the element category is a continuous thin landmark, acquiring the matching result comprises:
determining fold line representation information corresponding to the map element;
acquiring a plurality of candidate pixels by parsing the semantic element according to the fold line representation information;
performing a downsampling process on the plurality of candidate pixels to acquire a plurality of reference pixels; and
identifying a plurality of second map line segments corresponding to the plurality of reference pixels respectively from the map element according to the fold line representation information, wherein the plurality of reference pixels and the plurality of second map line segments together constitute the matching result;
wherein if the element category is a continuous thick landmark, acquiring the matching result comprises:
determining a second contour corresponding to the semantic element, the second contour comprising a plurality of second contour pixels;
determining a primary direction;
projecting each of the plurality of second contour pixels according to the primary direction to acquire a plurality of projection pixels; and
identifying a plurality of third map line segments corresponding to the plurality of projection pixels respectively from the map element, wherein the plurality of projection pixels and the plurality of third map line segments together constitute the matching result.