公开(公告)号：US20230351618A1

公开(公告)日：2023-11-02

申请号：US18338328

申请日：2023-06-20

Applicant: ZHEJIANG LAB

Inventor： Yechi MA ,  Wei HUA ,  Quan FENG ,  Shun ZHANG

IPC: G06T7/246

CPC classification number: G06T7/251 ,  G06T2207/10028 ,  G06T2207/20081 ,  G06T2207/20084

Abstract: A system and a method for detecting a moving target based on multi-frame point clouds. The system comprises a voxel feature extraction module; a transformer module used for matching and fusing the feature tensor sequence, fusing a first feature tensor with a second feature tensor, fusing the fused result with a third feature tensor, fusing the fused result with a fourth feature tensor, and repeating the fusing steps with a next feature tensor to obtain a final fused feature tensor; and an identification module used for extracting features from the final fused feature tensor and outputting detection information of a target. The method comprises the following steps: S1, constructing each system module; S2, training the model by the data in a training set; S3, predicting by the trained model.

公开(公告)号：US20230122343A1

公开(公告)日：2023-04-20

申请号：US17896055

申请日：2022-08-25

Applicant: ZHEJIANG LAB

Inventor： Wei HUA ,  Yechi MA ,  Shun ZHANG

IPC: G06V10/74 ,  G06V10/82

Abstract: An efficient across-camera target re-identification method based on similarity, which obtains a plurality of matching pairs and similarity scores thereof through two groups of targets to be matched; wherein for the matching pairs that are not matched by both parties, only a part of the matching pairs with higher similarity scores are selected each time, and the matching pairs are traversed according to the order of the similarity scores thereof from large to small, and the matching pairs and the similarity scores thereof are output as a matching result; when any target to be matched in a matching pair already appears in the matching result, the target cannot be output as the matching result; unmatched matching pairs are repeated traversed until the matching result reaches the expectation. The method firstly solves the multi-target matching problem based on similarity, and greatly reduces the time complexity and improves the efficiency.