Real-time multi-view detection of objects in multi-camera environments

Patent metadata

Field	Value
Publication number	US-2021256245-A1
Application number	US-201817251195-A
Country	US
Kind code	A1
Filing date	Sep 26, 2018
Priority date	Sep 26, 2018
Publication date	Aug 19, 2021
Grant date	—

Abstract

A mechanism is described for facilitating real-time multi-view detection of objects in multi-camera environments, according to one embodiment. A method of embodiments, as described herein, includes mapping first lines associated with objects to a ground plane; and forming clusters of second lines corresponding to the first lines such that an intersection point in a cluster represents a position of an object on the ground plane.

First claim

1 .- 19 . (canceled)

20 . An apparatus to facilitate real-time multi-view detection of objects in multi-camera environments, the apparatus comprising:
one or more processors to: map first lines associated with objects to a ground plane; and form clusters of second lines corresponding to the first lines such that an intersection point in a cluster represents a position of an object on the ground plane.

21 . The apparatus of claim 20 , wherein the one or more processors are further to generate the first lines to represent the objects such that a first line runs vertically from top to bottom of an object, wherein the first lines include vertical lines, and wherein the objects include one or more of persons, other living beings, and not living things.

22 . The apparatus of claim 20 , wherein the second lines intersection to generate clusters having intersection points, where each cluster represents an object group, and wherein the second lines includes principal lines.

23 . The apparatus of claim 20 , wherein the one or more processors are further to facilitate cameras to capture scenes having images of the objects, wherein a camera offers an input based on its capture of a scene of one or more of the objects, wherein the input represents a view of the one or more object from the camera's perspective.

24 . The apparatus of claim 20 , wherein the one or more processors are further to
detect whether the second lines include an isolated second line; if the isolated second line is detected, create a new object group to assign the isolated second line is assigned to the new object group; and if the isolated second line is not detected, assign the second lines to object groups.

25 . The apparatus of claim 20 , wherein the one or more processors are further to:
compute group centers for the object groups; and output positions of the objects within the group plane based on the group centers.

26 . The apparatus of claim 20 , wherein the one or more processors comprise one or more of a graphics processor and an application processor, wherein the graphics processor and the application processor are co-located on a common semiconductor package.

27 . A method for facilitating real-time multi-view detection of objects in multi-camera environments, the method comprising:
mapping first lines associated with objects to a ground plane; and forming clusters of second lines corresponding to the first lines such that an intersection point in a cluster represents a position of an object on the ground plane.

28 . The method of claim 27 , further comprising generating the first lines to represent the objects such that a first line runs vertically from top to bottom of an object, wherein the first lines include vertical lines, and wherein the objects include one or more of persons, other living beings, and not living things.

29 . The method of claim 27 , wherein the second lines intersection to generate clusters having intersection points, where each cluster represents an object group, and wherein the second lines includes principal lines.

30 . The method of claim 27 , further comprising facilitating cameras to capture scenes having images of the objects, wherein a camera offers an input based on its capture of a scene of one or more of the objects, wherein the input represents a view of the one or more object from the camera's perspective.

31 . The method of claim 27 , further comprising:
detecting whether the second lines include an isolated second line; if the isolated second line is detected, creating a new object group to assign the isolated second line is assigned to the new object group; and if the isolated second line is not detected, assigning the second lines to object groups.

32 . The method of claim 27 , further comprising:
compute group centers for the object groups; and output positions of the objects within the group plane based on the group centers.

33 . The method of claim 27 , wherein the method is facilitated by one or more processors comprising one or more of a graphics processor and an application processor, wherein the graphics processor and the application processor are co-located on a common semiconductor package.

34 . A computer-readable medium having stored thereon instructions which, when executed, cause a computing device to perform operations comprising:
mapping first lines associated with objects to a ground plane; and forming clusters of second lines corresponding to the first lines such that an intersection point in a cluster represents a position of an object on the ground plane.

35 . The computer-readable medium of claim 34 , wherein the operations further comprise generating the first lines to represent the objects such that a first line runs vertically from top to bottom of an object, wherein the first lines include vertical lines, and wherein the objects include one or more of persons, other living beings, and not living things.

36 . The computer-readable medium of claim 34 , wherein the second lines intersection to generate clusters having intersection points, where each cluster represents an object group, and wherein the second lines includes principal lines.

37 . The computer-readable medium of claim 34 , wherein the operations further comprise facilitating cameras to capture scenes having images of the objects, wherein a camera offers an input based on its capture of a scene of one or more of the objects, wherein the input represents a view of the one or more object from the camera's perspective.

38 . The computer-readable medium of claim 34 , wherein the operations further comprise:
detecting whether the second lines include an isolated second line; if the isolated second line is detected, creating a new object group to assign the isolated second line is assigned to the new object group; and if the isolated second line is not detected, assigning the second lines to object groups.

39 . The computer-readable medium of claim 34 , wherein the operations further comprise:
compute group centers for the object groups; and output positions of the objects within the group plane based on the group centers, wherein the method is facilitated by one or more processors comprising one or more of a graphics processor and an application processor, wherein the graphics processor and the application processor are co-located on a common semiconductor package.

Assignees

• Intel Corp

Inventors

• Li Qiang
• Tong Xiaofeng
• Fang Yikai
• Ling Chen
• Li Wenlong

Classifications

• G06T7/13 Primary

  Edge detection · CPC title

• G06V40/103

  Static body considered as a whole, e.g. static pedestrian or occupant recognition · CPC title

• G06V10/803

  of input or preprocessed data · CPC title

• G06V20/64

  Three-dimensional [3D] objects · CPC title

• H04N23/90

  Arrangement of cameras or camera modules, e.g. multiple cameras in TV studios or sports stadiums · CPC title