公开(公告)号：US20210374970A1

公开(公告)日：2021-12-02

申请号：US16888418

申请日：2020-05-29

Applicant: ZEBRA TECHNOLOGIES CORPORATION

Inventor： Duanfeng He ,  Vincent J. Daempfle

IPC: G06T7/194 ,  G06N3/08 ,  G06T7/55 ,  G06K9/00

Abstract: A method of generating neural network training image annotations includes training a first neural network to identify and segment hands in images using a first set of 2D images with hand portions segmented in each image; substantially simultaneously capturing both a second set of 2D images, and a third set of images including depth images, depicting hands holding a particular type of object; correlating each of the second set of images with corresponding images of the third set to identify and segment foregrounds from backgrounds in the second set of images; applying the trained first neural network to the identified foregrounds to identify hand portions of the foregrounds and segment object portions from identified hand portions; and training a second neural network, using the segmented object portions of the second set of images as training data, to identify the particular type of object in new images.

公开(公告)号：US20210334487A1

公开(公告)日：2021-10-28

申请号：US16856959

申请日：2020-04-23

Applicant: ZEBRA TECHNOLOGIES CORPORATION

Inventor： Sajan Wilfred ,  Christopher J. Fjellstad ,  Vincent J. Daempfle

IPC: G06K7/10 ,  G06K9/20 ,  G06K9/18 ,  G06K7/14

Abstract: Methods for accurate object tracking are disclosed herein. An example the method includes receiving, from a first optical imaging assembly having a first field of view (FOV), a first image captured over the first FOV and based on a decode of an indicia associated with an object of interest, identifying the object of interest within the first image. The method further includes determining a location of the object of interest within the first image and mapping the location of the object of interest within the first image to a predicted location of the object of interest within a second image, the second image being received from a second optical imaging assembly having a second FOV and the second image being captured over the second FOV.

公开(公告)号：US11335007B2

公开(公告)日：2022-05-17

申请号：US16888418

申请日：2020-05-29

Applicant: ZEBRA TECHNOLOGIES CORPORATION

Inventor： Duanfeng He ,  Vincent J. Daempfle

IPC: G06T7/55 ,  G06T7/194 ,  G06N3/08 ,  G06V40/10

Abstract: A method of generating neural network training image annotations includes training a first neural network to identify and segment hands in images using a first set of 2D images with hand portions segmented in each image; substantially simultaneously capturing both a second set of 2D images, and a third set of images including depth images, depicting hands holding a particular type of object; correlating each of the second set of images with corresponding images of the third set to identify and segment foregrounds from backgrounds in the second set of images; applying the trained first neural network to the identified foregrounds to identify hand portions of the foregrounds and segment object portions from identified hand portions; and training a second neural network, using the segmented object portions of the second set of images as training data, to identify the particular type of object in new images.

公开(公告)号：US11328140B2

公开(公告)日：2022-05-10

申请号：US16856959

申请日：2020-04-23

Applicant: ZEBRA TECHNOLOGIES CORPORATION

Inventor： Sajan Wilfred ,  Christopher J. Fjellstad ,  Vincent J. Daempfle

IPC: G06K7/10 ,  G06K7/14 ,  G06K9/18 ,  G06K9/20 ,  G06V10/22 ,  G06V30/224

Abstract: Methods for accurate object tracking are disclosed herein. An example the method includes receiving, from a first optical imaging assembly having a first field of view (FOV), a first image captured over the first FOV and based on a decode of an indicia associated with an object of interest, identifying the object of interest within the first image. The method further includes determining a location of the object of interest within the first image and mapping the location of the object of interest within the first image to a predicted location of the object of interest within a second image, the second image being received from a second optical imaging assembly having a second FOV and the second image being captured over the second FOV.