Granted Patent B2

US 12,361,700 · App. 18/236,747 · Granted Jul 15, 2025

Inventor: Hiroyoshi Miyano (Tokyo, JP)

Assignee: NEC CORPORATION

G06V10/82G06T7/215G06T7/251G06T7/70G06V10/28G06V20/52H04N7/18G06T2207/10016G06T2207/10024G06T2207/20076G06T2207/20081G06T2207/30196G06T2207/30232

Abstract

Claims (56)

1. An image processing system comprising:

at least one memory storing instructions; and

at least one processor configured to process the instructions to control the image processing system to:

receive an input of a plurality of image frames having different capturing times; and

perform one or more convolution calculations by using values of a background model of a neighboring region of one or more target pixels in the plurality of image frames to learn a parameter of a detection model for detecting a moving object, wherein

the plurality of background models include three background models, and these three background models are different from each other in the number of image frames to be processed;

the first background model has a largest number of image frames to be processed, the second background model has a second largest number of image frames to be processed, and the third background model has a smallest number of image frames to be processed;

the at least one processor configured to process the instructions to control the image processing system to:

calculate a first distance between the first background model and the second background model, a second distance between the second background model and the third background model, and a third distance the first background model and the third background model for each pixel or unit region of the plurality of image frames;

determine, based on the first distance, the second distance, and the third distance, whether or not the moving object is shown in a region corresponding to each pixel of the plurality of image frames; and

based on user input specifying whether some pixels in the plurality of image frames are pixels that correspond to the moving object or a pixel that does not show the moving object, learn parameters for detecting the moving object in the plurality of image frames.

2. The image processing system according to claim 1 , wherein

the at least one processor is configured to process the instructions to control the image processing system to:

learn, as the parameter, the one or more convolution calculations, and a threshold compared with a value obtained as a result of the one or more convolution calculations.

3. The image processing system according to claim 1 , wherein

the at least one processor is configured to process the instructions to control the image processing system to:

identify whether pixels that are selected at random from the plurality of image frames are a background region or an object region based on additional user inputs.

4. The image processing system according to claim 1 , wherein

the at least one processor is configured to process the instructions to control the image processing system to:

display an image frame including the moving object;

place a first icon on a background region of the image frame based on a first user input; and

place a second icon on an object region of the image frame based on a second user input.

5. An image processing method performed by at least one computer, the method comprising:

receiving an input of a plurality of image frames having different capturing times; and

performing one or more convolution calculations by using values of a background model of a neighboring region of one or more target pixels in the plurality of image frames to learn a parameter of a detection model for detecting a moving object, wherein

the plurality of background models include three background models, and these three background models are different from each other in the number of image frames to be processed;

the first background model has a largest number of image frames to be processed, the second background model has a second largest number of image frames to be processed, and the third background model has a smallest number of image frames to be processed;

the method comprises:

calculating a first distance between the first background model and the second background model, a second distance between the second background model and the third background model, and a third distance the first background model and the third background model for each pixel or unit region of the plurality of image frames;

determining, based on the first distance, the second distance, and the third distance, whether or not the moving object is shown in a region corresponding to each pixel of the plurality of image frames; and

based on user input specifying whether some pixels in the plurality of image frames are pixels that correspond to the moving object or a pixel that does not show the moving object, learning parameters for detecting the moving object in the plurality of image frames.

6. The image processing method according to claim 5 , wherein the method comprises:

learning, as the parameter, the one or more convolution calculations, and a threshold compared with a value obtained as a result of the one or more convolution calculations.

7. The image processing method according to claim 5 , wherein the method comprises:

identifying whether pixels that are selected at random from the plurality of image frames are a background region or an object region based on additional user inputs.

8. The image processing method according to claim 5 , wherein the method comprises:

displaying an image frame including the moving object;

placing a first icon on a background region of the image frame based on a first user input; and

placing a second icon on an object region of the image frame based on a second user input.

9. A non-transitory computer readable recording medium storing program instructions for causing a computer to perform:

receiving an input of a plurality of image frames having different capturing times; and

performing one or more convolution calculations by using values of a background model of a neighboring region of one or more target pixels in the plurality of image frames to learn a parameter of a detection model for detecting a moving object, wherein

the plurality of background models include three background models, and these three background models are different from each other in the number of image frames to be processed;

the first background model has a largest number of image frames to be processed, the second background model has a second largest number of image frames to be processed, and the third background model has a smallest number of image frames to be processed;

the program instructions causes the computer to perform:

calculating a first distance between the first background model and the second background model, a second distance between the second background model and the third background model, and a third distance the first background model and the third background model for each pixel or unit region of the plurality of image frames;

determining, based on the first distance, the second distance, and the third distance, whether or not the moving object is shown in a region corresponding to each pixel of the plurality of image frames; and

based on user input specifying whether some pixels in the plurality of image frames are pixels that reflect the moving object or a pixel that does not show the moving object, learning parameters for detecting the moving object in the plurality of image frames.

10. The non-transitory computer readable recording medium according to claim 9 , wherein the program instructions cause the computer to perform:

learning, as the parameter, the one or more convolution calculations, and a threshold compared with a value obtained as a result of the one or more convolution calculations.

11. The non-transitory computer readable recording medium according to claim 9 , wherein the program instructions cause the computer to perform:

identifying whether pixels that are selected at random from the plurality of image frames are a background region or an object region based on additional user inputs.

12. The non-transitory computer readable recording medium according to claim 9 , wherein the program instructions cause the computer to perform:

displaying an image frame including the moving object;

placing a first icon on a background region of the image frame based on a first user input; and

placing a second icon on an object region of the image frame based on a second user input.

Priority Claims (1)

Continuity (4)

References Cited (91)