Granted Patent B2

US 12,651,349 · App. 18/189,476 · Granted Jun 9, 2026

Inventors: Niko Nevatie (Helsinki, FI); Erkki Parkkulainen (Helsinki, FI)

Assignee: TGI SPORT VIRTUAL TECHNOLOGIES LIMITED

G06T7/194G06T7/13G06T7/60G06V10/25G06T2207/20081

Abstract

Claims (42)

1 . A computer-implemented training method for training a segmentation model to segment an image, the method comprising:

receiving a plurality of training data sets, the training data sets each comprising:

image data representing an image comprising a foreground and a background, the image data being captured by at least one visible electromagnetic radiation imaging device; and

background composition data defining a composition of at least a portion of the background of the image represented by the image data, wherein the background composition data defines a geometry of at least the portion of the background of the image, and wherein the background composition data defines a spatial location and/or orientation of the at least one visible electromagnetic radiation imaging device;

wherein the method comprises:

processing each training data set using the segmentation model to generate a candidate segmentation;

supplying the background composition data and the candidate segmentation to an error calculating algorithm to determine an error; and

updating the segmentation model in dependence on the determined error.

2 . A computer-implemented training method as claimed in claim 1 , wherein the background composition data defines a property of the background which is at least substantially constant.

3 . A computer-implemented training method as claimed in claim 1 , wherein the background comprises one or more advertising board, the background composition data defining a boundary and/or a profile of the one or more advertising board.

4 . A computer-implemented training method as claimed in claim 1 , wherein the image comprises at least one foreground area occurring in the foreground of the image, the segmentation model being configured to process the image data to segment the image to determine the at least one foreground area.

5 . A computer-implemented training method as claimed in claim 1 , wherein the image comprises at least one background area occurring in the background of the image, the segmentation model being configured to process the image data to segment the image to determine the at least one background area.

6 . A computer-implemented training method as claimed in claim 1 , wherein the image comprises at least one mixed area occurring in the foreground and the background of the image, the segmentation model being configured to process the image data to segment the image to determine the at least one mixed area.

7 . A non-transitory computer-readable medium having a set of instructions stored therein which, when executed, cause a processor to perform the method claimed in claim 1 .

8 . An image segmentation system for segmenting an image, the image segmentation system comprising one or more processors; wherein the one or more processors is configured to implement a segmentation model trained according to the method claimed in claim 1 .

9 . A content replacement system for replacing the content of an image, the content replacement system comprising an image segmentation system as claimed in claim 8 .

10 . An image processing system for processing an image, the image processing system comprising one or more processors having an electrical input for receiving image data captured by at least one visible electromagnetic radiation imaging device and representing a image comprising a foreground and a background;

wherein the one or more processors is configured to implement a segmentation model trained using the computer-implemented training method claimed in claim 1 ; the segmentation model being configured to segment the image to differentiate between the foreground and the background of the image.

11 . A computer-implemented method of processing a first image, the method comprising receiving image data captured by at least one visible electromagnetic radiation imaging device and representing an image comprising a foreground and a background;

wherein the method comprises implementing a segmentation model to segment the image to differentiate between the foreground and the background of the first image, the segmentation model being trained using the computer-implemented training method claimed in claim 1 .

12 . A system for training a segmentation model to segment an image; the system comprising:

at least one processor and at least one memory device, the at least one processor comprising at least one input configured to receive a plurality of training data sets, the training data sets each comprising:

image data representing an image comprising a foreground and a background, the image data being captured by at least one visible electromagnetic radiation imaging device; and

background composition data defining a composition of at least a portion of the background of the image represented by the image data, wherein the background composition data defines a geometry of at least the portion of the background of the image, and wherein the background composition data defines a spatial location and/or orientation of the at least one visible electromagnetic radiation imaging device;

wherein, in respect of each of the plurality of first training data sets, the at least one processor is configured to:

process each training data set using the segmentation model to generate a candidate segmentation;

supplying the background composition data and the candidate segmentation to an error calculating algorithm to determine an error; and

updating the segmentation model in dependence on the determined error.

13 . A system as claimed in claim 12 , wherein the background composition data defines a property of the background which is at least substantially constant.

14 . A system as claimed in claim 12 , wherein the background comprises one or more advertising board, the background composition data defining a boundary and/or a profile of the one or more advertising board.

15 . A system as claimed in claim 12 , wherein the image comprises at least one foreground area occurring in the foreground of the image, the segmentation model being configured to process the image data to segment the image to determine the at least one foreground area.

16 . A system as claimed in claim 12 , wherein the image comprises at least one background area occurring in the background of the image, the segmentation model being configured to process the image data to segment the image to determine the at least one background area.

17 . A system as claimed in claim 12 , wherein the image comprises at least one mixed area occurring in the foreground and the background of the image, the segmentation model being configured to process the image data to segment the image to determine the at least one mixed area.

18 . A computer-implemented training method for training a segmentation model to segment an image, the method comprising:

receiving a plurality of training data sets, the training data sets each comprising:

first image data representing a first image comprising a background; and

second image data representing a second image comprising at least a portion of the background of the first image;

wherein the method comprises processing each training data set using the segmentation model, the processing of each training data set comprising:

processing the first image data to generate background composition data defining a composition of at least a portion of the background of the image represented by the first image data, wherein the background composition data defines a geometry of at least the portion of the background of the image, and wherein the background composition data defines a spatial location and/or orientation of the at least one visible electromagnetic radiation imaging device;

segmenting the second image data to generate a candidate segmentation;

supplying the background composition data and the candidate segmentation to an error calculating algorithm and determining an error for the candidate segmentation; and

updating the segmentation model in dependence on the determined error.

Assignments (2)

Priority Claims (1)

Continuity (1)

References Cited (87)