Method, device, and program for deanonymizing a region containing a target

Abstract

In one aspect, a method is provided for de-anonymizing an object-containing region in an input image, the input image including a plurality of identified subjects, the method including: acquiring a historical image in which the object appears; identifying a region by comparing the input image with the acquired historical image to determine whether the object has interacted with at least one of the identified subjects; and de-anonymizing the region in which the object is included.

Description

[Technical Field]

[0001] The present invention relates to, but is not limited to, a method and apparatus for de-anonymizing a region containing an object and subjects that interact with it. [Background technology]

[0002] Video surveillance is widely used by law and security authorities as a primary tool for monitoring communities and preventing crime and terrorism. While the public supports its use due to its effectiveness in crime prevention, deterrence, and post-crime investigation, its intrusion into people's daily lives has raised privacy concerns about its misuse and abuse.

[0003] The captured video data contains private information, and for people who value their privacy and do not want their information to be made public, this indiscriminate filming is a serious problem.

[0004] Privacy violations can be a serious social problem. To address privacy violations, video surveillance needs to anonymize irrelevant and unnecessary data while maintaining its capabilities and performance. Summary of the Invention [Problem to be solved by the invention]

[0005] Existing facial recognition solutions frequently apply anonymization techniques to detected faces to protect data privacy. Similar to the limitations of all other object detection algorithms, the face detection algorithms in facial recognition solutions can face false negatives due to the algorithm's construction and training dataset. [Means for solving the problem]

[0006] A method is provided for de-anonymizing regions containing objects in an input image, the input image including a plurality of identified subjects, the method including: acquiring historical images in which the objects appear; identifying regions by comparing the input image with the acquired historical images to determine whether the objects have interacted with at least one of the identified subjects; and de-anonymizing the regions containing the objects.

[0007] An apparatus for de-anonymizing a region containing an object in an input image is provided, the apparatus having at least one memory element configured to store instructions and at least one processor configured to execute the instructions to acquire historical images in which the object appears, identify regions by comparing the input image with the acquired historical images to determine whether the object has interacted with at least one of the identified subjects, and de-anonymize the region containing the object. [Brief explanation of the drawings]

[0008] The accompanying drawings, in which like reference numerals indicate identical or functionally similar elements throughout the different views, and which, together with the following detailed description, are incorporated into and constitute a part of the specification, illustrate, by way of non-limiting example only, various embodiments and provide an explanation of various principles and advantages in accordance with the present embodiments.

[0009] Embodiments of the present invention will be better understood and readily apparent to those skilled in the art from the following written description, which is illustrated by way of example only and in conjunction with the drawings in which:

[0010] [Figure 1] FIG. 1 illustrates a system for de-anonymizing regions in an input image that include an object and identified subjects with which the object has interacted, in one embodiment of the present disclosure. [Figure 2A]FIG. 2A illustrates a prior art method for de-anonymizing a subject. [Figure 2B] FIG. 2B illustrates a method for de-anonymizing a target and identified subjects according to the present disclosure. [Figure 3] FIG. 3 illustrates a key advantage of one embodiment of the present disclosure. [Figure 4] FIG. 4 illustrates steps of how the present disclosure removes the background to generate a foreground image containing the identified subject, according to one embodiment of the present disclosure. [Figure 5] FIG. 5 shows that one of the identified subjects matches a person of interest according to one embodiment of the present disclosure. [Figure 6] FIG. 6 is a flowchart for searching for the appearance of a target person from a history image in one embodiment of the present disclosure. [Figure 7] FIG. 7 illustrates a flowchart for finding remaining identified subjects in a foreground image from the appearance of a target person in one embodiment of the present disclosure. [Figure 8] FIG. 8 illustrates a method for calculating coordinates of the appearance of a target subject and identified subjects in a historical image in one embodiment of the present disclosure. [Figure 9] FIG. 9 illustrates how historical co-occurrences are counted by intersection detection on historical images in one embodiment of the present disclosure. [Figure 10] FIG. 10 illustrates how the minimum and maximum X and Y coordinates are determined from the top left and bottom right points of a rectangle in one embodiment of the present disclosure. [Figure 11] FIG. 11 illustrates a flowchart for adaptively displaying target subjects and identified subjects (if any) in one embodiment of the present disclosure. [Figure 12] FIG. 12 shows another example of a method for performing interaction detection using three-dimensional distance calculation in one embodiment of the present disclosure. [Figure 13] FIG. 13 illustrates an exemplary computing device that may be used to perform the methods of the previous figures. DETAILED DESCRIPTION OF THE INVENTION

[0011] Disclosed herein are embodiments of an apparatus and method for de-anonymizing a region containing a subject that address one or more of the problems discussed above.

[0012] Furthermore, other desirable features and characteristics will become apparent from the following detailed description and the appended claims, taken in conjunction with the accompanying drawings and this background of this disclosure.

[0013] Terminology Subject - A subject is any suitable type of entity, including individuals, patients, and users.

[0014] The term subject or target is used herein to identify an individual, user, or patient of interest. A target subject is a subject selected by user input or a subject identified as being of interest.

[0015] Here, the term "subject" or "identified subject" is used to refer to a person associated with or appearing with the target subject in the input image (e.g., a partner or companion). For example, in the context of a pandemic outbreak, a subject is a person associated with or appearing with the target.

[0016] A user who is registered with the contact tracing server is called a registered user. A user who is not registered with the contact tracing server is called an unregistered user. A user can get a graphical representation of any subject on the network graph.

[0017] Contact Tracing Server - A contact tracing server is a server that hosts a software application program for receiving input, processing data, and objectively providing a graphical representation. The contact tracing server communicates with other servers (e.g., a remote assistance server) to manage requests. The contact tracing server communicates with the remote assistance server to display a graphical representation of potential subjects and target subjects. Contact tracing servers may use various protocols and procedures to manage data and provide the graphical representation.

[0018] Contact tracing servers are typically managed by a provider, which is an entity (e.g., a company or organization) that processes requests, manages the data, and displays contextual and customizable graphical representations. The server may include one or more computing devices used to process requests for graphical representations and provide contextual and customizable services.

[0019] Contact Tracing Account - A contact tracing account is a user's account that is registered with the contact tracing server. In some circumstances, a contact tracing account is not required for use with the remote assistance server. A contact tracing account contains user details (e.g., name, address, vehicle, etc.).

[0020] The contact tracing server manages the user's contact tracing account and the user's interactions with other external servers, along with the data exchanged.

[0021] Detailed Description Where reference is made in any one or more of the accompanying drawings to steps and / or features having the same reference number, these steps and / or features have the same function or operation for the purposes of this description, unless intended to the contrary.

[0022] It should be noted that the "Background" section and the above discussion of prior art relate to discussion of devices that form public knowledge through their use, and such should not be construed as a representation by the inventors or patent applicants that such devices in any way form part of the public knowledge in the art.

[0023] System 100

[0024] 1 shows a block diagram of a system 100 for de-anonymizing a region of a person of interest and regions of any identified subjects interacting with the person of interest in a historical image. The system 100 includes a requester device 102, a contact tracing server 108, a remote assistance server 140, remote assistance hosts 150A-150N, and sensors 142A-142N.

[0025] The requester device 102 communicates with the contact tracing server 108 and / or the remote assistance server 140 via connections 116 and 121, respectively. The connections 116 and 121 may be wireless (e.g., via NFC communication, Bluetooth, etc.) or over a network (e.g., the Internet). The connections 116 and 121 may also be network (e.g., the Internet) connections.

[0026] The contact tracing server 108 is further in communication with a remote assistance server 140 via a connection 120. The connection 120 may be a network (e.g., the Internet) connection. In some configurations, the contact tracing server 108 and the remote assistance server 140 are combined, and the connection 120 may be an interconnection bus.

[0027] The remote support server 140 communicates with the remote support hosts 150A to 150N via respective connections 122A to 122N, which may be a network (eg, the Internet).

[0028] The remote support hosts 150A to 150N are servers. Here, the term host is used to distinguish the remote support hosts 150A to 150N from the remote support server 140. The remote support hosts 150A to 150N are collectively referred to as the remote support hosts 150, and the term "remote support host 150" refers to any one of the remote support hosts 150. The remote support host 150 may be combined with the remote support server 140.

[0029] For example, the remote assistance host 150 may be managed by a hospital, and the remote assistance server 140 may be a central server that manages emergency calls and determines which remote assistance host 150 will transfer and acquire data such as image capture.

[0030] Sensors 142A-142N are connected to remote assistance server 140 or contact tracing server 108 via connections 144A-144N or 144A-144N, respectively. Herein, sensors 142A-142N are collectively referred to as sensors 142A-142N. Herein, connections 144A-144N are collectively referred to as connections 144, and connection 144 refers to any of connections 144. Similarly, connections 146A-146N are collectively referred to as connections 146, and connection 146 refers to any of connections 146. Connections 144 and 146 may be network connections. Sensor 142 may be any of an image capture device, a video capture device, and a motion sensor, and may be configured to transmit input corresponding to the type of sensor to at least one of contact tracing servers 108.

[0031] In an exemplary embodiment, each device 102 and 142 and server 108, 140, and 150 provides an interface to enable communication with other connected devices 102 and 142 and / or servers 108, 140, and 150. Such communication is facilitated by an application programming interface ("API"). Such an API may be part of a user interface that includes a programmatic interface such as a graphical user interface (GUI), a web-based interface, an application programming interface (API) and / or a set of remote procedure calls (RPCs) corresponding to interface elements, a messaging interface in which interface elements correspond to messages in a communication protocol, and / or any suitable combination thereof.

[0032] Use of the term "server" herein can refer to a single computing device or multiple interconnected computing devices that work together to perform a particular function, i.e., a server may be contained in a single hardware unit or may be distributed across multiple or many different hardware units.

[0033] Remote support server 140

[0034] The remote assistance server 140 is associated with an entity (e.g., a company, organization, or moderator of the service). In some configurations, the remote assistance server 140 is owned and operated by the entity that operates the server 108. In such configurations, the remote assistance server 140 may be implemented as part of the server 108 (e.g., a computer program module, a computing device, etc.).

[0035] The remote assistance server 140 may be configured to manage user registration. Registered users have a contact tracing account (see above) that contains the user's details. The registration procedure is called onboarding. A user can onboard to the remote assistance server 140 using any of the requester devices 102.

[0036] It is not necessary to have a contact tracing account on the remote assistance server 140 in order to access its features. However, some features are available to registered users, such as a graphical display of eligible subjects and potential subjects in other jurisdictions. These additional features are described below.

[0037] The user onboarding process is initiated by the user via one of the requester devices 102. In one configuration, the user downloads an application (including an API for interacting with the remote assistance server 140) to the sensor 142. In another configuration, the user accesses a website (including an API for interacting with the remote assistance server 140) on the requester device 102.

[0038] Registration details may include, for example, the user's name, the user's address, emergency contact information, or other important information, as well as the sensors 142 authorized to update the remote assistance account.

[0039] Once onboarded, users will have a contact tracing account with all their details stored.

[0040] Requester Device 102

[0041] A requester device 102 is associated with a subject (or request) that is the party to a contact tracing request initiated from the requester device 102. A requester may be an interested member of the public who assists in obtaining the data needed to obtain a graphical representation of a network graph. A requester device 102 may be a computing device such as a desktop computer, an interactive voice response (IVR) system, a smartphone, a laptop computer, a personal digital assistant computer (PDA), a mobile computer, a tablet computer, etc.

[0042] In one example configuration, requestor device 102 is a computing device in a watch or similar wearable, equipped with a wireless communication interface.

[0043] Contact Tracing Server 108

[0044] The contact tracing server 108 is as described in the terminology section above.

[0045] The contact tracing server 108 is configured to perform processing related to de-anonymizing regions of persons of interest and regions of any identified subjects in historical images.

[0046] Remote support host 150

[0047] The remote assistance host 150 is a server associated with an entity (eg, a company or organization) that manages (eg, establishes or manages) information about a subject.

[0048] In one configuration, the entities are hospitals. Thus, each entity has a remote support In some configurations, the remote host 150 is operated by the entity to manage resources. support The host 150 receives a warning signal that the subject of interest is likely to be a carrier of the virus.

[0049] Advantageously, such information is important to law enforcement agencies as it reduces the time spent reviewing camera footage and investigating connections between persons of interest who may have interacted in the past.

[0050] That information is especially useful during a pandemic, allowing building management and health departments to conduct contact tracing more efficiently and effectively. The network of interactions between a subject and other subjects helps identify how a disease spreads from one person to another.

[0051] Traditionally, related subjects who may have interacted with the subject are not displayed, i.e., only the subject is de-anonymized. Other conventional techniques use detection techniques for anonymization, i.e., all detected people in the input image are anonymized. Existing techniques use foreground or background extraction to anonymize moving subjects. More commonly, the background of the image can be blurred or image anonymization can be performed on predefined segments / regions of the image, or vice versa, to exclude regions of interest for image anonymization.

[0052] The present disclosure provides image anonymization that de-anonymizes a person of interest and subjects who may have interacted with that person while protecting uninvolved persons. While combining a detected face algorithm of a person of interest with anonymization of the full image background helps to better protect the privacy of uninvolved persons, it may also hide details of the interaction between the person of interest and their potential associates in their surroundings.

[0053] Sensor 142

[0054] The sensors 142 are associated with a user associated with the requester device 102. Details of how the sensors may be utilized are provided below.

[0055] 2A shows how the conventional technology de-anonymizes a subject. As shown in FIG. 2A, the conventional technology de-anonymizes subject 202. Subject 204 is a subject that has interacted with subject 202, and according to the conventional technology, subject 204 is not de-anonymized.

[0056] 2B illustrates how the present disclosure de-identifies the subject and the identified subject. As shown in FIG. 2B, the present disclosure de-anonymizes the subject 202 and the subject 204, which is a subject that may have interacted with the subject 202. More information can be provided below.

[0057] Figure 3 illustrates a key advantage of one embodiment of the present disclosure. An input image 302 may be received, including a target 202 and a subject 204 among multiple subjects. At 304, image anonymization is performed based on a set or predefined algorithm. At step 306, pre-stored and labeled historical images of all target persons are retrieved from a database. At step 308, a processing step is performed on the retrieved historical information to detect interactions between the target and other subjects by calculating the intersection between the rectangle coordinates of the two persons. At step 310, regions in the input image 302 that include the target and subjects that have interacted with the target are de-anonymized. An output image 312 is output.

[0058] FIG. 4 illustrates the steps of how the present disclosure de-anonymizes targets and identified subjects, according to one embodiment of the present disclosure.

[0059] 4, an input image 402 is received. The method includes applying background image subtraction to the input image 402 for foreground image extraction, as shown at 404. A sample of the extracted foreground image is shown at 406.

[0060] Figure 5 may follow Figure 4. Facial recognition may be performed on the extracted foreground image for a predetermined list of subjects, as shown at 502. As a result of the facial recognition, #4 may match a subject found in the subject list.

[0061] FIG. 6 may follow FIG. 5. In FIG. 、6 All previously stored historical images of the matched object are retrieved, as shown in 602. That is, historical information about the object is retrieved and images in which the object appears are identified, as shown in 604.

[0062] FIG. 7 may follow FIG. 6. In FIG. 7, subjects (e.g., #1, #2, and #3) that appear with the subject in the input image are identified. It should be understood that in some embodiments, this may be only one subject. At 702, face matching of each identified subject is performed on the image acquired for the subject (#4). Subjects #1 and #3 from the input image are detected from the historical information acquired for subject #4. In the historical information acquired for subject #4, #1 appears twice, and #3 appears once in the historical information acquired for subject #4. Therefore, the number of co-occurrences between #4 and #1 is 2, the number of co-occurrences between #4 and #2 is 0, and the number of co-occurrences between #4 and #3 is 1.

[0063] Figure 8 may follow Figure 7. Based on the historical information shown in 802, the count of interactions between #4 and #1 is 2, the count of interactions between #4 and #2 is 0, and the count of interactions between #4 and #3 is 1.

[0064] At 804, the pixel-wise rectangular coordinates of each subject are calculated by image segmentation, where X = minimum of X (position of left edge), Y = minimum of Y (position of top edge), H = maximum height of the image segment, and W = maximum weight of the image segment. In one embodiment, the rectangular coordinates may be calculated based on XY1 = {X, Y}, XY2 = {X + W, Y}, XY3 = {X, Y + H}, and XY4 = {X + W, Y + H}.

[0065] At 806, the rectangle coordinates of #4 and #3 are calculated. In one embodiment, the rectangle coordinates of #4 are XY1#4={950,300}, XY2#4={1100,300}, XY3#4={950,600}, and XY4#4={1100,600}. Additionally or alternatively, the rectangle coordinates of #3 are XY1#3={650,300}, XY2#3={780,300}, XY3#3={650,610}, and XY4#3={780,610}.

[0066] FIG. 9 may be a continuation of FIG. 8. As described in FIG. 7, face matching for each identified subject is performed on the image acquired for the subject (#4). Subjects #1 and #3 from the input image are detected from the historical information acquired for the subject #4. In the historical information acquired for the subject #4, #1 appears twice, and #3 appears once in the historical information acquired for the subject #4. Therefore, #4 and #1 co-occur twice, #4 and #2 co-occur zero times, and #4 and #3 co-occur once. Image segmentation may be performed on each historical image, and it may be determined that the rectangular coordinates of #4 and #3 intersect, as shown in 902. In this example, #3 and #4 may be de-anonymized in the input image.

[0067] Figure 10 may be a continuation of Figure 9. Figure 10 shows in more detail how rectangle coordinates are used to determine if there is an intersection between a subject and a target subject. The following may be used:

number

number

[0068] An intersection is detected when the values ​​of Intersect-W and Intersect-H are both greater than 0. The rectangular coordinates of A, B, and C are shown in 1002: A1 {100,100}, A2 {200,250}, B1 {20,200}, B2 {100,240}, C1 {250,90}, and C2 {360,260}.

[0069] Based on equations (1) and (2), an intersection is detected between A(R1) and B(R2). Intersect-W = Math.max(0, Math.min(200, 120) - Math.max(100, 40)) = Math.max(0, 120 - 100) = 20 Intersect-H = Math.max(0, Math.min(250, 240) - Math.max(100, 200)) = Math.max(0, 240 - 200) = 40

[0070] Based on equations (1) and (2), no intersection is detected between A(R1) and C(R2). Intersect-W = Math.max(0, Math.min(200, 360) - Math.max(100, 250)) = Math.max(0, 200 - 250) = 0 Intersect-H = Math.max(0, Math.min(250, 260) - Math.max(100, 90)) = Math.max(0, 250 - 100) = 150

[0071] 11 illustrates a method for de-anonymizing regions containing objects in one embodiment of the present disclosure. The method begins in step 1102. In step 1104, an input image containing objects and subjects is received. In one embodiment, in step 1106, a pre-defined camera background is obtained. In step 1108, background subtraction is performed on the input image received in 1102 to obtain a foreground image. In step 1110, face recognition is performed on the foreground image after background subtraction. In step 1112, detected faces are matched against a pre-defined list of objects.

[0072] At step 1114, it is determined whether the person of interest is found. If it is determined that the person of interest is not found, then all subjects in the input image are anonymized at step 1116. If it is determined that the person of interest is found, then at step 1118, all pre-stored historical images for the matched subject of interest are retrieved.

[0073] In step 1120, it is determined whether a historical image is found. If it is determined that a historical image is not found, in step 1132, all subjects are anonymized and any found persons of interest are revealed for the input image. If it is determined that a historical image is found, in step 1122, face matching is performed on all faces detected in the input image for all captured historical images.

[0074] At step 1124, it is determined whether an interacting person is found. If it is determined that an interacting person is not found, then at step 1132, processing is performed to anonymize all subjects and reveal the person of interest in the input image. If it is determined that an interacting person is found, then at step 1126, intersection detection is performed on all co-occurrences with the found subject of interest identified in the input image.

[0075] In step 1128, a list of detected persons of interest that intersected with at least one person of interest is obtained. In step 1130, all subjects are anonymized to reveal the persons of interest in the input image and a list of persons that intersected in the input image. The method then proceeds to the end.

[0076] 12 illustrates how intersection between a subject and a target subject is determined in one embodiment of the present disclosure. In FIG. 12, the three-dimensional (3D) distance between two people is obtained by converting the people in a two-dimensional (2D) image to their real-world 3D positions, and then calculating the distance between the two people to determine whether it is within a predetermined distance (e.g., 1 meter).

[0077] Assuming the inputs f=35mm, sensor=35mm, height=170cm, the distance from the camera shown in 1202 to subject B is calculated as follows: (170cm * 35mm / (80p / 720p) * 35mm) = 5950 / 3.8888 = 1,530.0350 cm

[0078] The distance from the camera shown in 1202 to subject A is calculated as follows: (170cm * 35mm / (100p / 720p) * 35mm) = 5950 / 4.8611 = 1,224.0028 cm

[0079] Thus, the distances of A and B from the camera shown in 1204 are calculated as follows: abs((distance A from camera)-(distance B from camera)) = abs(2,040.0466 - 1,530.0350) = 510.0116 cm = ~510 cm

[0080] The distance between A and B shown in 1204 is calculated as follows: √ (204 * 204 + 510 * 510) = √(41,616 + 260,100) = √(301,716) = 549.2868 cm = ~5.49 meters

[0081] Based on the calculated distance between A and B from the camera, it is determined whether subjects A and B have intersected in the image.

[0082] One or more computer devices 1300 (hereinafter also referred to as "computer system 1300") may be used to implement the methods described above. An exemplary computer device 1300 may be used to implement the system 100 shown in Figure 1. The following description of the computer device 1300 is merely exemplary and is not intended to be limiting.

[0083] 13, the example computing device 1300 includes a processor 1307 for executing software routines. While a single processor is shown for clarity, the computing device 1300 may also include a multi-processor system. The processor 1307 is connected to a communications infrastructure 1306 for communicating with other components of the computing device 1300. The communications infrastructure 1306 may include, for example, a communications bus, crossbar, or network.

[0084] The computing device 1300 further includes a main memory 1308, such as random access memory (RAM), and a secondary memory 1310. The secondary memory 1310 may include a storage drive 1312, which may be, for example, a hard disk drive, a solid-state drive, or a hybrid drive, and / or a removable storage drive 1317, which may include a magnetic tape drive, an optical disk drive, a solid-state storage drive (such as a USB flash drive, a flash memory device, a solid-state drive, a memory card, etc.). The removable storage drive 1317 reads from and / or writes to a removable storage medium 1377 in a well-known manner. The removable storage medium 1377 may include a magnetic tape, an optical disk, a non-volatile memory storage medium, etc., which is read from and written to by the removable storage drive 1317. As will be appreciated by those skilled in the art, the removable storage medium 1377 includes a computer-readable storage medium having computer-executable program code instructions and / or data stored thereon.

[0085] In alternative implementations, the secondary memory 1310 may additionally or alternatively include other similar means for allowing computer programs or other instructions to be loaded into the computer system 1300. Such means may include, for example, a removable storage device 1322 and an interface 1350. Examples of removable storage devices 1322 and interfaces 1350 include program cartridges and cartridge interfaces (e.g., those found in video game console devices), removable memory chips (e.g., EPROM or PROM) and associated sockets, removable solid-state storage drives (such as USB flash drives, flash memory devices, solid-state drives, memory cards, etc.), and other removable storage devices 1322 and interfaces 1350 that allow software and data to be transferred from the removable storage device 1322 to the computer system 1300.

[0086] The computing device 1300 also includes at least one communications interface 1327. The communications interface 1327 allows software and data to be transferred between the computing device 1300 and external devices via the communications interface 1327. In various embodiments of the present invention, the communications interface 1327 allows data to be transferred between the computing device 1300 and a data communications network, such as a public or private data communications network. The communications interface 1327 can be used to exchange data between different computing devices 600, where such computing device 1300 forms part of an interconnected computer network. Examples of communications interfaces 1327 can include a modem, a network interface (e.g., an Ethernet card), a communications port (e.g., serial, parallel, printer, GPIB, IEEE 1394, RJ45, USB, etc.), an antenna with associated circuitry, etc. The communications interface 1327 can be wired or wireless. The software and data transferred via the communications interface 1327 are in the form of signals, which can be electronic, electromagnetic, optical, or other signals receivable by the communications interface 1327. These signals are provided to the communications interface via communications path 1327.

[0087] As shown in FIG. 13, the computing device 1300 further includes a display interface 1302 that operates to render images on an associated display 1350 and an audio interface 1352 that operates to play audio content through associated speakers 1357.

[0088] As used herein, the term "computer program product" may refer, in part, to removable storage medium 1377, removable storage device 1322, a hard disk attached to storage drive 1312, or software carrying a carrier wave over communications path 1327 (wireless link or cable) to communications interface 1327. A computer-readable storage medium refers to any non-transitory, non-volatile, tangible storage medium that provides recorded instructions and / or data to computing device 1300 for execution and / or processing. Examples of such storage media include magnetic tape, CD-ROM, DVD, Blu-ray® disk, hard disk drive, ROM or integrated circuit, solid-state storage drive (such as a USB flash drive, flash memory device, solid-state drive, memory card), hybrid drive, magneto-optical disk, or computer-readable card such as a PCMCIA card, whether such device is internal or external to computing device 1300. Examples of transitory or non-tangible computer-readable transmission media that may also be involved in providing software, application programs, instructions and / or data to computing device 1300 include wireless or infrared transmission channels, network connections to other computers or networked devices, and the Internet or intranet, including email transmissions and information stored on websites and the like.

[0089] Computer programs (also referred to as computer program code) are stored in main memory 1308 and / or secondary memory 1310. Computer programs may also be received via communications interface 1327. When such computer programs are executed, they enable computing device 1300 to perform one or more features of the embodiments described herein. In various embodiments, when the computer programs are executed, processor 1307 can perform the features of the above-described embodiments. Thus, such computer programs represent controllers of computing system 1300.

[0090] The software may be stored on a computer program product and loaded into the computer device 1300 using the removable storage drive 1317, the storage drive 1312, or the interface 1350. The computer program product may be a non-transitory computer-readable medium. Alternatively, the computer program product may be downloaded to the computer system 1300 via communications path 1327. The software, when executed by the processor 1307, causes the computer device 1300 to perform the operations necessary to implement the methods as described above.

[0091] It should be understood that the embodiment of Figure 13 is presented merely as an example to illustrate the operation and structure of system 100. Accordingly, in some embodiments, one or more features of computing device 1300 may be omitted. Also, in some embodiments, one or more features of computing device 1300 may be combined together. Furthermore, in some embodiments, one or more features of computing device 1300 may be split into one or more component parts.

[0092] It will be understood by those skilled in the art that various changes and / or modifications may be made to the present invention as illustrated in the specific embodiments without departing from the spirit or scope of the invention as broadly described, and the present embodiments are therefore considered in all respects to be illustrative and not restrictive.

[0093] This application claims the benefit of priority to Singapore Patent Application No. 10202108597X, filed on August 5, 2021, the disclosure of which is incorporated herein by reference in its entirety.

Claims

1. A method for detecting a plurality of identified subjects, comprising: obtaining an input image including a plurality of identified subjects; generating an output image from the input image by anonymizing each of the subjects; acquiring one or more historical images in which the object appears; By comparing the input image with each of the history images, for each of the plurality of subjects, calculate the number of times that both the target and the subject are included in the history image, and determine whether each of the subjects has interacted with the target based on the number of times; and, if the object is determined to have interacted with at least one of the subjects, de-anonymizing in the output image an area displaying the object and an area displaying the subject determined to have interacted with the object.

2. The method described in claim 1, wherein the target and each of the subjects are identified from each other by performing facial matching.

3. Obtaining an input image including a plurality of identified subjects; generating an output image from the input image by anonymizing each of the subjects; acquiring one or more historical images in which the object appears; By comparing the input image with each of the history images, for each of the plurality of subjects, calculate the number of times that both the target and the subject are included in the history image, and determine whether each of the subjects has interacted with the target based on the number of times; and when it is determined that the object has interacted with at least one of the subjects, de-anonymizing in the output image an area displaying the object and an area displaying the subject determined to have interacted with the object.

4. Obtaining an input image including a plurality of identified subjects; generating an output image from the input image by anonymizing each of the subjects; acquiring one or more historical images in which the object appears; By comparing the input image with each of the history images, for each of the plurality of subjects, calculate the number of times that both the target and the subject are included in the history image, and determine whether each of the subjects has interacted with the target based on the number of times; A program that causes a computer to, when it is determined that the target has interacted with at least one of the subjects, de-anonymize an area in the output image that displays the target and an area that displays the subject that has been determined to have interacted with the target.

Images