US 12,243,311 B2
Methods and systems for vehicle verification
Taolue Xue, Beijing (CN); Tianming Zhang, Beijing (CN); Zhiheng Wang, Beijing (CN); Shudong Wang, Beijing (CN); Jie Li, Beijing (CN); Hui Meng, Beijing (CN); Tianyu Chen, Hangzhou (CN); Guiting Dai, Hangzhou (CN); Zhaohui Wu, Hangzhou (CN); and Duoqing Zhou, Hangzhou (CN)
Assigned to BEIJING DIDI INFINITY TECHNOLOGY AND DEVELOPMENT CO., LTD., Beijing (CN)
Filed by BEIJING DIDI INFINITY TECHNOLOGY AND DEVELOPMENT CO., LTD., Beijing (CN)
Filed on Jun. 23, 2022, as Appl. No. 17/808,532.
Application 17/808,532 is a continuation of application No. PCT/CN2020/138862, filed on Dec. 24, 2020.
Claims priority of application No. 201911349534.4 (CN), filed on Dec. 24, 2019.
Prior Publication US 2022/0335722 A1, Oct. 20, 2022
Int. Cl. G06K 9/00 (2022.01); G06V 10/82 (2022.01); G06V 20/50 (2022.01)
CPC G06V 20/50 (2022.01) [G06V 10/82 (2022.01); G06V 2201/08 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method, implemented on a computing device including at least one processor and at least one storage medium, the at least one storage medium including a set of instructions for vehicle verification, the computing device being communicatively connected to a client related to a vehicle to be verified via a network, the method comprising:
sending, to the client via the network, at least one vehicle verification instruction for instructing a user of the client to perform at least one verification action on the vehicle;
receiving, from the client via the network, vehicle verification information related to the vehicle, wherein the vehicle verification information includes a plurality of images acquired by the client via an imaging device; and
for each image of the plurality of images, determining reference information of one or more preset points related to the vehicle in the image by processing the image using at least one model, and determining a vehicle state corresponding to the image based on the reference information of the one or more preset points, the at least one model being trained machine learning model;
determining, based on the plurality of vehicle states corresponding to the plurality of images, at least one vehicle action related to the vehicle; and determining a verification result indicating whether the vehicle is a real vehicle by determining whether the at least one vehicle action is consistent with the at least one verification action.