Granted Patent B2

US 12,530,776 · App. 18/189,467 · Granted Jan 20, 2026

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

Assignee: TGI SPORT VIRTUAL TECHNOLOGIES LIMITED

G06T7/11G06T7/194G06T2207/20081G06T2207/30196G06T2207/30221

Abstract

Claims (51)

1 . A computer-implemented training method for training a segmentation model to segment an image, the method comprising receiving a plurality of first training data sets, the first training data sets each comprising:

first image data representing a first image consisting of a background, the first image data being captured by at least one visible electromagnetic radiation imaging device; and

first annotation data identifying the presence solely of the background in the first image;

wherein the method comprises, for each first training data set:

processing the first image data automatically to generate a background composition data set defining a composition of at least a portion of the background of the first image represented by the first image data, the processing of the first image data comprising identifying background features having at least one of a fixed shape and a fixed profile;

processing the first image data using the segmentation model to generate a first candidate segmentation, the segmentation model using the background composition data when segmenting the first image;

supplying the first annotation data to an error calculating algorithm to determine a first error for the first candidate segmentation; and

updating the segmentation model in dependence on the determined first error.

2 . A computer-implemented training method as claimed in claim 1 , wherein the method comprises updating the segmentation model to generate a first updated segmentation model.

3 . A computer-implemented training method as claimed in claim 1 , wherein the method comprises generating the first annotation data by processing first segmentation data for differentiating between the foreground and the background of the first image.

4 . A computer-implemented training method as claimed in claim 3 , wherein the first segmentation data is captured by at least one non-visible electromagnetic radiation imaging device at at least substantially the same time as the first image data is captured by the at least one visible electromagnetic radiation imaging device.

5 . A computer-implemented training method as claimed in claim 1 , wherein the method comprises receiving a plurality of second training data sets, the second training data sets each comprising:

second image data representing a second image comprising a foreground and a background, the second image data being captured by at least one visible light imaging device.

6 . A computer-implemented training method as claimed in claim 5 , wherein the second training data sets each comprise second segmentation data for differentiating between the foreground and the background of the second image;

wherein the method comprises, for each second training data set:

generating a second target segmentation in dependence on the second segmentation data;

processing the second image data set using the segmentation model to generate a second candidate segmentation;

comparing the second candidate segmentation and the second target segmentation to determine a second error between the second candidate segmentation and the second target segmentation; and

updating the segmentation model in dependence on the determined second error.

7 . A computer-implemented training method as claimed in claim 6 , wherein the method comprises updating the segmentation model in dependence on the second training output to generate a second updated segmentation model.

8 . A computer-implemented training method as claimed in claim 3 , wherein the first segmentation data is captured by at least one non-visible electromagnetic radiation imaging device.

9 . 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 .

10 . 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 .

11 . A content replacement system for replacing the content of an image, the content replacement system comprising an image segmentation system as claimed in claim 10 .

12 . A system for training a segmentation model to segment an image into one or more segments, 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 first training data sets, the first training data sets each comprising:

first image data representing a first image consisting of a background or of a foreground, the first image data being captured by at least one visible electromagnetic radiation imaging device; and

first annotation data identifying the presence solely of the background or the foreground in the first image;

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

process the first image data automatically to generate a background composition data set defining a composition of at least a portion of the background of the first image represented by the first image data, the processing of the first image data comprising identifying background features having at least one of a fixed shape and a fixed profile;

process the first image data using the segmentation model to segment the first image to generate a first candidate segmentation, the segmentation model using the background composition data when segmenting the first image;

supply the first annotation data to an error calculating algorithm to determine a first error for the first candidate segmentation; and

update the segmentation model in dependence on the determined first error.

13 . A system as claimed in claim 12 , wherein the at least one processor is configured to update the segmentation model to generate a first updated segmentation model.

14 . A system as claimed in claim 12 , wherein the at least one processor is configured to generate the first annotation data by processing first segmentation data for differentiating between the foreground and the background of the first image.

15 . A system as claimed in claim 14 , wherein the first segmentation data is captured by at least one non-visible electromagnetic radiation imaging device at at least substantially the same time as the first image data is captured by the at least one visible electromagnetic radiation imaging device.

16 . A system as claimed in claim 12 , wherein the at least one processor is configured to receive a plurality of second training data sets, the second training data sets each comprising:

second image data representing a second image comprising a foreground and a background, the second image data being captured by at least one visible light imaging device.

17 . A system as claimed in claim 16 , wherein the second training data sets each comprise second segmentation data for differentiating between the foreground and the background of the second image;

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

process the second image data set using the segmentation model to generate a second candidate segmentation;

compare the second candidate segmentation and the second target segmentation to determine a second error between the second candidate segmentation and the second target segmentation; and

update the segmentation model in dependence on the determined second error.

18 . A system as claimed in claim 17 , wherein the at least one processor is configured to update the segmentation model in dependence on the second training output to generate a second updated segmentation model.

19 . A system as claimed in claim 12 , wherein the first segmentation data is captured by at least one non-visible electromagnetic radiation imaging device.

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

wherein the one or more processors is configured to:

process the first image data automatically to generate a background composition data set defining a composition of at least a portion of the background of the first image represented by the first image data, the processing of the first image data comprising identifying background features having at least one of a fixed shape and a fixed profile;

implement a segmentation model, the segmentation model being configured to segment the first image to differentiate between the foreground and the background of the first image;

wherein the segmentation model uses the background composition data when segmenting the first image.

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

wherein the method comprises implementing a segmentation model to segment the first 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 .

Assignments (2)

Priority Claims (1)

Continuity (1)

References Cited (80)