Method and apparatus for performing privacy masking by reflecting characteristic information of objects

Patent metadata

Field	Value
Publication number	US-10671877-B2
Application number	US-201916239775-A
Country	US
Kind code	B2
Filing date	Jan 4, 2019
Priority date	Mar 9, 2018
Publication date	Jun 2, 2020
Grant date	Jun 2, 2020

Abstract

An image masking method is provided. The method includes: extracting an object from an image; obtaining characteristic information about the extracted object by analyzing the extracted object; determining whether the extracted object is a masking target according to an input setting value or the obtained characteristic information; and performing masking such that the obtained characteristic information is reflected on the extracted object, in response to determining that the extracted object is the masking target among a plurality of objects extracted from the input image, wherein the setting value is set by an input designating at least a partial region in the input image, and wherein in the determining whether the extracted object is the masking target, an object positioned in the at least a partial region is determined as the masking target among the extracted objects.

First claim

What is claimed is:

1. An image masking method of a surveillance system in which a processor and a memory are included and an operation is controlled by the processor, the image masking method comprising:
extracting an object from an input image;
obtaining characteristic information about the extracted object by analyzing the extracted object;
determining whether the extracted object is a masking target according to an input setting value and the obtained characteristic information; and
performing masking on the extracted object based on the obtained characteristic information, in response to determining that the extracted object is the masking target among a plurality of objects extracted from the input image,
wherein the input setting value is set by an input designating a partial region in the input image,
wherein in the determining whether the extracted object is the masking target, when a first object among the extracted objects is positioned in the partial region, the first object is determined as the masking target, and when a second object among the extracted objects is positioned outside the partial region, the second object is not determined as the masking target, and
wherein the obtained characteristic information which is reflected on the extracted object is a characteristic information for monitoring the extracted object.

2. The image masking method of claim 1 , wherein the obtained characteristic information comprises information about at least one of color, shape, motion, identity, position, and size.

3. The image masking method of claim 1 , wherein the performing the masking comprises:
generating a masking object by reflecting the obtained characteristic information; and
performing the masking on the extracted object determined as the masking target using the masking object.

4. The image masking method of claim 3 , wherein the masking object is changed according to a change in the obtained characteristic information.

5. The image masking method of claim 3 , wherein the masking object is an avatar on which the obtained characteristic information is reflected.

6. The image masking method of claim 5 , wherein at least one of a face, a hair, a top, a bottom, a gender, and an age of the avatar is determined according to the obtained characteristic information.

7. The image masking method of claim 3 , wherein the masking object is a silhouette on which the obtained characteristic information is reflected.

8. The image masking method of claim 1 , wherein the setting value is further set by an input designating at least one object among the extracted objects, and
wherein in the determining whether the extracted object is the masking target, the designated at least one object is determined as the masking target among the extracted objects.

9. The image masking method of claim 1 , wherein the setting value comprises an input designating the at least a partial region in the input image, and an input designating sensitivity with respect to the designated at least a partial region for detecting a dynamic object,
wherein the sensitivity with respect to the at least a partial region is set to be higher than sensitivity with respect to other regions in the input image to more extensively detect the dynamic object, and
wherein the determining whether the extracted object is the masking target comprises determining a dynamic object satisfying the sensitivity with respect to the designated at least a partial region as the masking target among objects positioned in the designated at least a partial region.

10. The image masking method of claim 1 , wherein in the determining whether the extracted object is the masking target, an object having characteristic information satisfies the input setting value is determined as the masking target among the extracted objects.

11. The image masking method of claim 1 , wherein in the determining whether the extracted object is the masking target, only an object having characteristic information satisfying the setting value is determined as the masking target among a same type of objects.

12. An image masking method of a surveillance system in which a processor and a memory are included and an operation is controlled by the processor, the image masking method comprising:
extracting an object from an input image;
obtaining characteristic information about the extracted object by analyzing the extracted object;
determining whether the extracted object is a masking target according to an input setting value and the obtained characteristic information; and
performing masking on the extracted object based on the setting value and obtained characteristic information, in response to determining that the extracted object is the masking target among a plurality of objects extracted from the input image,
wherein in the determining whether the extracted object is the masking target, only an extracted object having characteristic information satisfying the setting value is determined as the masking target among a same type of objects,
wherein the obtained characteristic information which is reflected on the extracted object is a characteristic information for monitoring the extracted object,
wherein the input setting value is further set b an input designating a partial region on the image, and
wherein in the determining whether the extracted object is the masking target, when a first object among the extracted objects is positioned in the partial region, the first object is determined as the masking target, and when a second object among the extracted objects is positioned outside the partial region, the second object is not determined as the masking target.

13. The image masking method of claim 12 , wherein the performing the masking comprises:
generating a masking object by reflecting the obtained characteristic information on the extracted object; and
performing the masking on the extracted object determined as the masking target using the generated masking object.

14. The image masking method of claim 13 , wherein the masking object is changed according to a change in the obtained characteristic information about the object.

15. The image masking method of claim 14 , wherein the masking object is an avatar on which the obtained characteristic information is reflected, and at least one of a face, a hair, a top, a bottom, a gender, and an age of the avatar is determined according to the obtained characteristic information.

16. The image masking method of claim 12 , wherein the setting value comprises an input designating at least a partial region in the input image, and an input designating sensitivity with respect to the designated at least a partial region for detecting a dynamic object,
wherein the sensitivity with respect to the at least a partial region is set to be higher than sensitivity with respect to other regions in the input image to more extensively detect the dynamic object, and
wherein the determining whether the extracted object is the masking target comprises determining a dynamic object satisfying the sensitivity with respect to the designated at least a partial region as the masking target among objects positioned in the designated at least a partial region.

17. An image masking method of a surveillance system in which a processor and a memory are included and an operation is controlled by the processor, the image masking method comprising:
extracting an object from an input image;
obtaining characteristic information about the extracted object by analyzing the extracted object;
determining whether the ext

Assignees

• Hanwha Techwin Co Ltd

Inventors

• Choi Jin Hyuk
• Jeong Song Taek
• Song Jae Cheon

Classifications

• G06T2213/12

  Rule based animation · CPC title

• G06T13/40

  of characters, e.g. humans, animals or virtual beings · CPC title

• G06T7/90

  Determination of colour characteristics · CPC title

• G06K9/46 Primary

  Physics · mapped topic

• G06F21/60

  Protecting data · CPC title