KR-20260067159-A - Method and Apparatus for Masking Surveillance Video for Privacy Protection

KR20260067159AKR 20260067159 AKR20260067159 AKR 20260067159AKR-20260067159-A

Abstract

The embodiments relate to a masking method and apparatus for protecting the privacy of surveillance images, comprising: a step of automatically detecting a face region within an image; a step of extracting the detected face region in a size that is a multiple of 128-bit; a step of masking the extracted region by performing block encryption using a symmetric key method; a step of storing the size and median coordinates of the face region as metadata for each frame of the masked image; and a step of restoring the masked region to its original state using the stored metadata in response to a request from an authenticated user.

Inventors

주현수
박민철
고현우
임성일

Assignees

한국과학기술연구원

Dates

Publication Date: 20260512
Application Date: 20241105

Claims (10)

In a masking method for privacy protection of surveillance images performed by a processor, Step of automatically detecting face regions within an image; A step of extracting the detected face region in a size that is a multiple of 128-bit; A step of masking the extracted region by performing symmetric key block encryption; A step of storing the size and median coordinates of the face region as metadata for each frame of the masked image; and A masking method for privacy protection of surveillance video, characterized by including the step of restoring the masked area to its original state using the stored metadata in accordance with the request of an authenticated user.
In paragraph 1, A masking method for privacy protection of surveillance images, characterized in that the above metadata includes the xy size and median coordinates (α, β) of the face region.
In paragraph 1, A masking method for privacy protection of surveillance video, characterized in that the above restoration step provides the symmetric key only to the authenticated user to decrypt the masked area.
In paragraph 1, A masking method for privacy protection of surveillance video, characterized by further including the step of partially unmasking only specific frames of the masked video.
In paragraph 1, A masking method for protecting the privacy of surveillance images, characterized in that the above block encryption processes the extracted face region using a standardized symmetric key block encryption method to perform masking without damaging the original image.
In paragraph 1, A masking method for privacy protection of surveillance images, characterized in that the detection step is performed using a face recognition-based face region tracking algorithm.
In paragraph 1, A masking method for privacy protection of surveillance images, characterized by simultaneously detecting and masking multiple face regions for each frame of the above image.
In paragraph 1, A masking method for privacy protection of surveillance video, characterized by tracking the location of a masked area using the stored metadata and performing accurate restoration at that location.
In paragraph 1, A masking method for protecting the privacy of surveillance images, characterized by storing the original image and the masked image as a single image instead of storing them separately when storing the masked image.
In a masking device for privacy protection of surveillance footage, Face detection unit that automatically detects face regions within an image; A region extraction unit that extracts the above-described face region in a size that is a multiple of 128-bit; An encryption unit that performs symmetric key block encryption on the extracted region to mask it; A storage unit that stores the size and median coordinates of the face region as metadata for each frame of the masked image; and A masking device for privacy protection of surveillance video, characterized by including a restoration unit that restores the masked area to its original state using the stored metadata in accordance with a request from an authenticated user.

Description

Method and Apparatus for Masking Surveillance Video for Privacy Protection The present invention relates to a masking method and device for protecting the privacy of surveillance footage, and more specifically, to a masking method and device for preventing the exposure of privacy-related information, such as a person's face, when monitoring footage transmitted from a surveillance camera. [Explanation of government-supported R&D] This study was conducted with the support of the National Police Agency's field support technology development project for efficient policing activities (establishment of an augmented reality (XR)-based complex terror response education and training testbed, project number: PR08-04-000-21). Recently, the installation of surveillance cameras, such as CCTVs, has been rapidly increasing to ensure urban safety and prevent crime. However, the footage collected by these cameras contains personal information of an unspecified number of individuals, raising concerns about privacy infringement. As a solution to this, the conventional method was a privacy masking technique that involved covering a specific pre-set area before transmitting video from a surveillance camera. This aimed to resolve the issue of privacy infringement by obscuring privacy-related information when the transmitted video was played back at a security control center or similar facility. However, when masking video encoded using the latest video encoding standards, while privacy information can be hidden, other video information is also compromised. This is because recent video codecs encode data by cross-referencing it to increase compression ratios. Consequently, even if only a portion corresponding to the masked area is modified, distortion occurs in areas outside the selected region during decoding. In addition, the reason for masking only privacy-sensitive parts of the video instead of encrypting the entire video is to protect privacy while still allowing non-sensitive information in the video to be utilized. However, if unintended areas of the video are damaged, it does not meet the original purpose of masking. For example, in a video surveillance system where only a pedestrian's face is obscured while other information is visible to the control center to identify crimes or other accidents, due to the characteristics of recent video codecs, there may be cases where other video information is compromised even if only specific areas are encrypted, making it impossible to obtain the necessary information. Therefore, there is a need for a new masking technology that can safely protect the privacy information of the video while preserving other video information without damage. To more clearly explain the technical solution of the embodiments of the present invention or the prior art, the drawings necessary for the description of the embodiments are briefly introduced below. It should be understood that the drawings below are for the purpose of explaining the embodiments of this specification only and are not for the purpose of limitation. Additionally, for clarity of explanation, some elements in the drawings below may be depicted with various modifications, such as exaggeration or omission. Figure 1 is a conceptual diagram illustrating the problems of the face masking/unmasking method for protecting the privacy of unspecified individuals in existing video surveillance systems. FIG. 2 is a flowchart of a masking method for privacy protection of surveillance images according to an embodiment of the present invention. FIG. 3 is a block diagram of a masking device for privacy protection of surveillance images according to one embodiment of the present invention. FIGS. 4a to 4c are examples illustrating a detailed method of facial masking through the encryption of an image portion according to an embodiment of the present invention. FIGS. 5A and 5B are examples illustrating a detailed method for unmasking a face through decoding a masked image portion of FIG. 4C according to an embodiment of the present invention. FIG. 6 is an example of applying a masking method and device for privacy protection of surveillance images according to an embodiment of the present invention to an image management and transmission system. The terms used in this specification have been selected to be as widely used and general as possible, taking into account their functions; however, these may vary depending on the intent or convention of those skilled in the art or the emergence of new technologies. Additionally, in specific cases, terms have been selected at the applicant's discretion, and in such cases, their meanings will be described in the relevant explanatory sections of the specification. Therefore, it should be noted that the terms used in this specification should be interpreted based on their actual meaning and the overall content of this specification, rather than merely their names. Additionally, the embodiments described herein may