Masking detected objects in a video stream

Abstract

Disclosed is a method for masking a detected object in a video stream. A camera is arranged in a camera system including a camera and at least another device, and for the another device, a position and a field of view are known, and for the camera, a position and a field of view are known. In addition, the field of view of the another device and the field of view of the camera are non-overlapping. Information indicating that an object is approaching the field of view of the camera is obtained in the camera. The obtained information is determined based on information from the another device indicating the position and direction of movement of the object and the known position and field of view of the camera and the another device. In response to the obtained information, a threshold for detecting an object to be masked in a video stream captured by the camera is lowered. The object to be masked in the video stream is detected using the lowered threshold, and masking of the detected object in the video stream is inserted.

Description

Technical Field

[0001] The present invention relates to the masking of objects in a video stream captured by a camera, and in particular to the masking of objects when the objects first enter the field of view of the camera. Background Art

[0002] While video surveillance of a particular scene can improve overall security, it may be desirable to prevent certain objects in the scene from being identifiable, for example, by a person viewing a recorded video clip from a camera capturing the scene (such as a surveillance camera). For example, for privacy reasons, it may be desirable that the identity of a person captured by a camera, or details such as a vehicle's license plate, should not be directly derivable simply by viewing the recorded video clip. Such protection of certain objects can be achieved by masking the objects in the video stream before outputting the video stream to, for example, a display or storage device. Such masking may include, for example, overlaying the object with a solid color, blurring the object, pixelating the object, or even making the object more or less transparent in each image of the output video stream.

[0003] Before a mask for a specific object can be inserted, the location of the object within the image must first be estimated. This can be achieved by using an object detection algorithm that has been trained / configured to detect a specific class of objects within the image (such as faces, people, license plates, etc.). Once the location of the object has been detected, a mask can be applied to the object.

[0004] However, it is important to mask the object in all images of the video stream, even in situations where it is difficult to detect the object. For example, when the object first enters the camera's field of view, the object detection algorithm may not correctly detect the object because only a portion of the object is initially in the camera's field of view. This can cause the object to fail to be masked. Summary of the Invention

[0005] In order to at least partially address the above-identified problem of unreliable masking of objects when the objects first enter the field of view of a camera, the present disclosure provides an improved method of masking detected objects in a video stream captured by a camera, a corresponding camera and a non-transitory computer-readable storage medium as defined in the accompanying independent claims.

[0006] According to a first aspect, a method for masking a detected object in a video stream captured by a camera is provided. The camera is arranged in a camera system including a camera and at least another device. For the another device, the position and field of view are known, and for the camera, the position and field of view are known, and the field of view of the another device and the field of view of the camera are non-overlapping. In the camera, information indicating that an object is approaching the field of view of the camera is obtained. The obtained information is determined based on information indicating the position and movement direction of the object from the another device and the known positions and fields of view of the camera and the further device. In response to the obtained information, a threshold for detecting an object to be masked in the video stream captured by the camera is lowered. Then, the object to be masked in the video stream is detected using the lowered threshold, and a mask of the detected object is inserted in the video stream.

[0007] Because the position and field of view of the further device, as well as the position and field of view of the camera, are known, information indicating that an object is approaching the camera's field of view can be received by the camera based on the information indicating the object's position and direction of movement from the further device. By using this received information to lower the threshold for detecting an object to be blocked in the video stream captured by the camera, the probability of detecting the object when it first enters the camera's field of view is increased.

[0008] The further device may be, for example, a sensor.

[0009] The threshold may be a threshold for what is detected as foreground and what is detected as background. Thus, for detection of an object based on detection of the foreground, the probability of having detected the object when it first enters the field of view of the camera is increased.

[0010] The threshold value may specify a confidence level above which the object is determined to belong to the class of objects to be blocked in the video stream. Thus, object detection is based on the confidence level of the object belonging to the class of objects to be blocked in the video stream, increasing the probability that the object is detected when it first enters the camera's field of view.

[0011] The obtained information can further indicate a sub-region of the camera's field of view in which the expected object first appears. In a lowering operation, the threshold is then lowered only in the indicated sub-region. This increases the probability of detecting the object when it first enters the camera's field of view, while reducing the risk of falsely detecting the object in regions other than the sub-region.

[0012] This information can further indicate the time at which the object will first appear in the camera's field of view. Knowing the estimated time at which the object will appear in the camera's field of view allows for better timing in lowering the threshold. This reduces the risk of falsely detecting an object before it enters the camera's field of view.

[0013] The obtained information from the further device may further indicate that the object is shielded by the further device.

[0014] The obtained information from the further device may further indicate an object class of the object.

[0015] The obtained information from the further device may further indicate a confidence level related to the object category.

[0016] In the act of lowering, the threshold may be lowered in areas along the periphery of the field of view, thereby increasing the probability of detecting an object when it first enters the camera's field of view, while reducing the risk of falsely detecting an object in areas other than the periphery of the field of view.

[0017] The further device may be one of a further camera, an IR camera, a thermal camera, a radar, a sonar and a lidar.

[0018] The method may further include increasing the threshold after detecting an object in the camera's field of view. If the object is then detected using the increased threshold, maintaining masking of the detected object in the video stream. If the object is not detected using the increased threshold, discontinuing masking of the detected object in the video stream. This reduces the risk of falsely detecting an object after the object has fully entered the camera's field of view.

[0019] The action of raising the threshold value can be performed at a predetermined time after the object is detected. This enables reducing the risk of falsely detecting an object after the object has completely entered the field of view of the camera.

[0020] According to a second aspect, a non-transitory computer-readable storage medium is provided having stored thereon instructions for implementing the method according to the first aspect when executed in a camera having at least a processor.

[0021] According to a third aspect, a camera for masking a detected object in a video stream captured by a camera is provided. The camera includes circuitry configured to perform an acquisition function, a reduction function, a detection function, and an insertion function when the camera is arranged in a camera system including a camera and at least another device, wherein the position and field of view of the another device are known, and the position and field of view of the camera are known, and wherein the field of view of the another device and the field of view of the camera are non-overlapping. The acquisition function is configured to obtain information indicating that an object is approaching the field of view of the camera, wherein the obtained information is determined based on information indicating the position and direction of movement of the object from the another device and the known positions and fields of view of the another device and the camera. The reduction function is configured to, in response to the obtained information, reduce a threshold for detecting an object to be masked in the video stream captured by the camera. The detection function is configured to detect the object to be masked in the video stream using the reduced threshold, and the insertion function is configured to insert a mask of the detected moving object in the video stream.

[0022] Where applicable, the above-mentioned optional additional features of the method according to the first aspect also apply to the camera according to the third aspect. To avoid unnecessary repetitions, reference is made to the above.

[0023] Further applicability of the present invention will become apparent from the detailed description given below. However, it should be understood that although preferred embodiments of the present invention are indicated, various changes and modifications within the scope of the present invention will become apparent to those skilled in the art from the detailed description, and therefore the detailed description and specific examples are given by way of illustration only.

[0024] Therefore, it is to be understood that, because the specific component parts of the described devices or the actions of the described methods can be changed, the present invention is not limited to such devices and methods. It is also to be understood that the terms used herein are only used for the purpose of describing specific embodiments and are not intended to be limiting. It must be noted that, as used in the specification and the appended claims, the words "a", "an", "the" and "said" are intended to mean that there are one or more elements, unless the context clearly indicates otherwise. Thus, for example, a reference to a "unit" or "the unit" may include several devices, etc. In addition, the words "comprise", "include", "contain" and similar expressions do not exclude other elements or steps. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The above and other aspects of the present invention will now be described in more detail with reference to the accompanying drawings, which are not to be considered limiting but are for illustration and understanding.

[0026] Figure 1 A schematic diagram is shown relating to an embodiment of a camera for masking a detected object in a video stream captured by the camera.

[0027] Figure 2 A flow chart is shown relating to an embodiment of a method for masking a detected object in a video stream captured by a camera. DETAILED DESCRIPTION

[0028] The present invention will now be described hereinafter with reference to the accompanying drawings in which currently preferred embodiments of the invention are illustrated. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein.

[0029] Figure 1 A schematic block diagram of an embodiment of a camera 100 for masking detected objects in a video stream captured by the camera 100 is shown. The camera 100 can be any type of camera, such as a surveillance camera, a security camera, or a monitoring camera, that can be arranged in a camera system that includes the camera 100 and at least one further device, wherein the position and field of view of the camera 100 are known, and the position and field of view of the further device are also known. The further device can be, for example, another camera. The camera 100 captures a sequence of image frames (e.g., in the form of a video), wherein each image frame depicts a scene defined by the field of view of the camera 100.

[0030] The camera 100 includes circuitry 130 and may include a receiver 110. The camera 100 will also typically include an encoder (not shown), an image sensor (not shown), and an image processor (not shown).

[0031] The circuit 130 is configured to perform the functions of the camera 100. The circuit 130 may include a processor 132, such as a central processing unit (CPU), a microcontroller, or a microprocessor. The processor 132 is configured to execute program code. The program code may, for example, be configured to perform the functions of the camera 100.

[0032] The camera 100 may further include a memory 140. The memory 140 may be one or more of a cache, a flash memory, a hard drive, a removable medium, a volatile memory, a non-volatile memory, a random access memory (RAM), or another suitable device. In a typical arrangement, the memory 140 may include non-volatile memory for long-term data storage and volatile memory used as system memory for the circuit 130. The memory 140 may exchange data with the circuit 130 over a data bus. There may also be accompanying control lines and an address bus between the memory 140 and the circuit 130.

[0033] The functionality of the camera 100 may be implemented in the form of executable logic routines (e.g., lines of code, software programs, etc.) stored on a non-transitory computer-readable medium (e.g., memory 140) of the camera 100 and executed by the circuitry 130 (e.g., using the processor 132). Furthermore, the functionality of the camera 100 may be a stand-alone software application or form part of a software application that performs additional tasks associated with the camera 100. The described functionality may be considered a method that the processing unit (e.g., the processor 132 of the circuitry 130) is configured to perform. Furthermore, while the described functionality may be implemented in software, such functionality may also be performed via dedicated hardware or firmware, or some combination of hardware, firmware, and / or software.

[0034] The circuitry 130 of the camera 100 is configured to perform an acquisition function 141 , a reduction function 142 , a detection function 143 , and an insertion function.

[0035] The obtaining function 141 is configured to obtain information indicating that an object is approaching the field of view of the camera, wherein the obtained information is determined based on information indicating the position and movement direction of the object from the further device and the known positions and fields of view of the further device and the camera. The obtained information can be received in the receiver 110.

[0036] The lowering function 142 is configured to lower a threshold for detecting an object to be masked in a video stream captured by the camera in response to the obtained information.

[0037] The detection function 143 is configured to detect objects to be masked in the video stream using a lowered threshold.

[0038] The insertion function 144 is configured to insert a mask of a detected moving object in the video stream.

[0039] The masking contemplated herein may be, for example, a solid / opaque color, may be semi-transparent, may include applying motion blur to the object so that the object is no longer easily identifiable, and / or, for example, applying forced pixelation and / or blurring of the object in the output image stream so that the object is no longer easily identifiable, etc. In other contemplated embodiments, the masking may include making the object itself at least partially transparent in the output image stream so that the background is visible through the object. This may be possible if, for example, a background image is available (e.g., from an earlier time instance) without a previous image of the object.

[0040] In this context, detection can be achieved using one or more commonly used object detection algorithms available in various fields of computer technology, such as computer vision and / or image processing. Such algorithms can include both non-neural network methods and neural network methods. However, the minimum requirement is that the algorithm (or combination of algorithms) used is able to determine, at least under ideal conditions, whether a specific object (such as a face, body, license plate, etc.) is present in the image that is considered to be masked. As long as the above requirements are met, it does not matter whether the algorithm used is, for example, feature-based, template-based, and / or motion-based. For example, a detector can be implemented using one or more neural networks specifically trained for this purpose. For the purposes of this disclosure, it is also assumed that the algorithm used may have difficulty correctly detecting objects when they first enter the field of view of the camera 100, because at that time they are initially only partially within the field of view of the camera 100.

[0041] Camera 100 is configured to be arranged in a camera system including camera 100 and at least one further device. Specifically, the camera system including camera 100 and the further device should be arranged and configured so that the position and field of view of the further device, as well as the position and field of view of camera 100, are known, and so that information indicating the position and direction of movement of an object can be provided from the further device. Information indicating that an object is approaching the field of view of camera 100, to be received by camera 100, can then be prepared in a centralized solution using a central server, or in a distributed solution using the further device and camera 100. In a centralized solution, only the central server needs to know the position and field of view of the further device, as well as the position and field of view of camera 100. After receiving the information indicating the position and direction of movement of the object from the further device, the central server can then prepare information indicating that the object is approaching the field of view of camera 100 and provide this information to camera 100. In a distributed solution, the further device may send information indicating the position and direction of movement of the object directly to the camera 100, and the camera 100 may then determine, knowing its own position and field of view, whether the object is approaching the field of view of the camera 100, and trigger one or more of the three actions when it is approaching. In the latter case, the information indicating that the object is approaching the field of view of the camera 100 will be the information indicating the position and direction of movement of the object received from the further device.

[0042] To determine the position and direction of movement of an object, another device is typically required to provide at least a two-dimensional representation of the object over time. Thus, the further device may be, but is not limited to, another camera, an IR camera, a thermal camera, radar, lidar, or sonar. In this case, in a centralized solution, the further device can determine the relative position and direction of movement relative to its own position and field of view by detecting the object in a sequence of image frames. A central server can then, knowing the absolute or relative positions and fields of view of both the further device and camera 100, determine whether the object is approaching the field of view of camera 100 and, if so, provide this information to camera 100. In a distributed solution, knowledge of the position and field of view of the further device and camera 100 relative to a common coordinate system can be located in the further device and camera 100, respectively. Alternatively, each of the further device and camera 100 may know its relative position relative to the other. The position and direction of movement of the object can then be determined in the further device by detecting the object in a sequence of image frames, and information indicating the object's position and direction of movement can be provided to camera 100. Then, knowing its own position and field of view, camera 100 can determine whether an object is approaching its field of view and, if so, trigger one or more of three actions. In the latter case, the information indicating that the object is approaching the field of view of camera 100 will be information indicating the object's position and direction of movement received from another device. Based on the detection of the object by the other device in the sequence of image frames, it is possible to determine an estimated time when the object will enter the field of view of camera 100. Furthermore, based on the position and direction of movement, it is possible to determine the sub-area of ​​the field of view of camera 100 in which the object will appear.

[0043] Generally speaking, it is not important how the position and field of view of the further device and the position and field of view of camera 100 are determined, as long as they are determined centrally or distributedly and subsequently known, such that camera 100 receives information indicating that an object is approaching its field of view, where the information is based on information indicating the position and direction of movement of the object from the further device, the position and field of view of the further device, and the position and field of view of camera 100. For example, the position and field of view of the further device and the position and field of view of camera 100 can be determined when the camera system is installed, for example, using GPS and / or a map-based or manually input location and orientation of the installation into the system connected to the camera system. Alternatively, the position and field of view of the further device and the position and field of view of camera 100 can be determined using a self-learning system that receives information about the detection and movement of objects in the further device and the camera, and determines over time whether the position and field of view of the further device and the position and field of view of the second device and / or whether an object having a particular position and direction of movement based on the further device is approaching the field of view of camera 100.

[0044] When applicable, the following Figure 2 The optional additional features of the method 200 described also apply to Figure 1 A camera 100 is described.

[0045] Figure 2 1 shows a flow chart relating to an embodiment of a method 200 for masking a detected object in a video stream captured by a camera. Figure 1 A camera system comprising a camera and at least a further device is described in more detail. For the further device, the position and field of view are known, and for the camera, the position and field of view are known. The method 200 comprises obtaining S210 in the camera information indicating that an object is approaching the field of view of the camera, wherein the obtained information is determined based on information indicating the position and movement direction of the object from the further device and the known positions and fields of view of the camera and the further device. Figure 1 This is described in more detail. The method 200 further comprises: in response to the obtained information, lowering S220 a threshold for detecting an object to be masked in a video stream captured by the camera; detecting S230 the object to be masked in the video stream using the lowered threshold; and inserting S240 a mask of the detected object in the video stream.

[0046] By using information from another device indicating the position and direction of movement of an object, as well as the known positions and fields of view of the camera and the further device, to obtain information indicating that an object is approaching the field of view of the camera, the camera will know that the object is about to appear in its field of view before the object actually appears. In addition, when the object first appears in the field of view of the camera, only a portion of the object will typically be visible in one or more images of the video stream captured by the camera. Depending on the type of detection used, the object may not be detected until all or at least a considerable portion of the object is within the field of view of the camera. Therefore, in order to increase the probability of the object being detected even when only a portion of the object is within the field of view of the camera when the object first appears in the field of view, the sensitivity of the detection is increased based on the information obtained. Specifically, this is accomplished by lowering the threshold used in S220 for detecting objects to be blocked in the video stream captured by the camera.

[0047] Because lowering the threshold for detecting objects to be masked in the video stream captured by the camera increases sensitivity in all areas of the camera's field of view, the obtained information can further indicate a sub-region within the camera's field of view where the expected object first appears. The S220 threshold is then lowered only in the indicated sub-region. In all other areas of the camera's field of view, the threshold is not lowered. This increases detection sensitivity only in the sub-region where the expected object first appears in the image, thereby reducing the risk of false detection and masking of objects in other areas of the field of view. Additionally or alternatively, the S220 threshold can be lowered in areas along the periphery of the field of view. This achieves a similar effect, since objects close to the camera's field of view will first appear somewhere along the periphery.

[0048] Before the object appears in the field of view of the camera, information that the object is approaching the field of view of the camera is obtained S210 in the camera.

[0049] The obtained information may further indicate a point in time when the object will appear in the field of view of the camera. Then, starting from the indicated time point, the threshold is lowered ( S220 ). This will increase the sensitivity of detection starting from the indicated time point and thus reduce the risk of false detection and masking of objects before the indicated time point.

[0050] Information obtained from another device (e.g., another camera) may further indicate that the object is blocked by the further device. The indication that the object is blocked by the further device may increase the probability that the object should be blocked by the camera, and thus, the threshold for detecting the object to be blocked may be further lowered.

[0051] The information obtained from the camera of another device may further indicate the object category of the object. The camera can then use this information to determine whether the object should also be shielded by the camera. If not, the action of reducing S220 can be overruled.

[0052] The information obtained from the camera of the further device may further indicate a confidence level related to the object category. If the indication of the further device has a high confidence level for the object category, and the category is a category that should be blocked in the camera, the threshold for detecting the object to be blocked may be further lowered for the object category.

[0053] In an embodiment, the threshold value may be a threshold value for what is detected as the foreground and what is detected as the background. For example, the background model may define for each of a plurality of spatial regions in an image frame of a video stream whether the spatial region belongs to the background or the foreground in the video sequence. Under the condition that the image data of the spatial region of the image frame has not changed significantly relative to the image data of the spatial region of the preceding image frame in the video stream before a predetermined time limit, the spatial region may be determined to belong to the background, and otherwise determined to belong to the foreground. Then, lowering the threshold value for what is detected as the foreground and what is detected as the background may be to lower the amount of change that is considered to be a significant change, or to increase the time limit within which the spatial region that is considered to belong to the background should not have changed significantly.

[0054] In other embodiments, the threshold specifies a confidence level above which the object is determined to belong to the class of objects to be masked in the video stream. Then, the threshold is lowered ( S220 ) to reduce the confidence level required to determine that the object belongs to the class of objects to be masked. For example, if all faces should be masked, the threshold for identifying faces is lowered, for example, from 80% to 30%, so that masking is performed for objects that are faces with a 30% probability.

[0055] The method may further include raising the threshold again (S250) after detecting the object (S230) using the lowered threshold. If the object is detected using the raised threshold, then maintaining (S260) the masking of the detected object in the video stream. If the object is not detected using the raised threshold, then discontinuing (S270) the masking of the detected object in the video stream. For example, the threshold (S250) may be raised at a predetermined time after detecting (S230) the object. The predetermined time may be selected so that it is likely that the entire object is within the field of view of the camera.

[0056] When applicable, Figure 1 The above optional additional features described for the camera 100 also apply to Figure 2 Method 200 is described.

[0057] In the alternative, provide the same Figure 2 A method for masking a detected object in a video stream captured by a camera is similar to the method 200 described above. The camera is arranged in a manner similar to the method 200 described above. Figure 1The described camera system is similar to a camera system comprising a camera and at least one further device. For the further device, the position and field of view are known, and for the camera, the position and field of view are known. This alternative method differs in that some feature of the camera system exists such that detection of an object by the further device is sufficient to determine that the object is approaching the camera's field of view. For example, this may be due to the camera and the further device being arranged in a particular manner relative to each other. For example, the further device may be arranged in a corridor or along a road so that it has a field of view covering a portion of the corridor or road that is in front of the portion covered by the camera's field of view. This means that an object moving along the road or corridor and detected by the further device is approaching the camera's field of view and is likely to enter the camera's field of view within a period of time that depends on the distance between the further device and the camera and the speed of the detected object. The alternative method includes obtaining ( S210 ) information in the camera indicating that an object is approaching the camera's field of view, wherein the obtained information is determined based on information from the further device indicating that the object has been detected and the recognition that the object detected by the further device indicates that the object is approaching the camera's field of view. The alternative method further comprises: lowering a threshold for detecting an object to be masked in a video stream captured by the camera in response to the obtained information; detecting S230 the object to be masked in the video stream using the lowered threshold; and inserting S240 a mask of the detected object in the video stream.

[0058] By using information from yet another device indicating that an object has been detected, and recognizing that the object detected by yet another device indicates that the object is approaching the camera's field of view, the camera will be aware of the object's impending appearance before the object actually appears in its field of view. Furthermore, when an object first appears in the camera's field of view, typically only a portion of the object will be visible in one or more images of the video stream captured by the camera. Depending on the type of detection used, the object may not be detected until all, or at least a substantial portion, of the object is within the camera's field of view. Therefore, in order to increase the probability of detecting an object even when only a portion of the object is within the camera's field of view when the object first appears in the field of view, the sensitivity of the detection is increased based on the information obtained. Specifically, this is accomplished by lowering the threshold used to detect objects to be blocked in the video stream captured by the camera.

[0059] Because lowering the threshold for detecting objects to be blocked in the video stream captured by a camera increases sensitivity across all areas of the camera's field of view, the information obtained can further indicate the sub-area of ​​the camera's field of view where the expected object first appears. This information can be obtained based on a priori knowledge of where the object will first appear after being detected by another device. For example, if the other device is positioned along a road or corridor ahead of the camera, the object detected by the other device will appear in a sub-area of ​​the camera that includes the portion of the road or corridor closest to the field of view of the other device. The threshold is then lowered only in this indicated sub-area. The threshold is not lowered in all other areas of the camera's field of view. This increases detection sensitivity only in the sub-area of ​​the image where the expected object first appears, thereby reducing the risk of false detection and blocking of objects in other areas of the field of view. Additionally or alternatively, the threshold can be lowered in areas along the periphery of the field of view. This achieves a similar effect, since objects close to the camera's field of view will first appear somewhere along the periphery.

[0060] Before the object appears in the field of view of the camera, information that the object is approaching the field of view of the camera is obtained in the camera.

[0061] The information obtained can further indicate the point in time when the object will appear in the camera's field of view. The threshold is then lowered starting from the indicated time point. This increases the sensitivity of detection starting from the indicated time point and thus reduces the risk of false detection and masking of objects before the indicated time point.

[0062] When applicable, Figure 1 The above optional additional features described for the camera 100 also apply to Figure 2 Describes the alternative method.

[0063] It will be appreciated by those skilled in the art that the present invention is not limited to the foregoing embodiments. Rather, numerous modifications and variations are possible within the scope of the appended claims. By studying the drawings, the disclosure, and the appended claims, a skilled artisan can understand and implement such modifications and variations in practicing the claimed invention.

Claims

1. A method for masking a detected object in a video stream captured by a camera, wherein: The camera is arranged in a camera system comprising the camera and at least a further device, wherein for the further device a position and a field of view are known, and for the camera a position and a field of view are known, and wherein the field of view of the further device and the field of view of the camera are non-overlapping, the method comprising: In the camera, information indicating that an object is approaching the field of view of the camera is obtained, wherein the obtained information is determined based on information indicating the position and movement direction of the object from the further device and the known positions and fields of view of the camera and the further device; In response to the information obtained, lowering a threshold for detecting an object to be masked in the video stream captured by the camera, wherein the threshold specifies a confidence level above which an object is determined to belong to a class of objects to be masked in the video stream, or wherein the threshold is a threshold for what is detected as a foreground and what is detected as a background, and specifies an amount of change in image data in an image frame relative to image data in a previous image frame, above which an object is determined to belong to the foreground; detecting an object to be masked in the video stream using the reduced threshold; and A detected mask of the object is inserted into the video stream.

2. The method according to claim 1, wherein: The obtained information further indicates a sub-region of the field of view of the camera in which the object is expected to first appear, and wherein, in the act of lowering, the threshold is lowered only in the indicated sub-region.

3. The method according to claim 1, wherein: The obtained information further indicates a point in time when the object will first appear in the field of view of the camera, and wherein, starting from the indicated point in time, the threshold is lowered.

4. The method according to claim 1, wherein: The information obtained from the further device further indicates that the object is shielded by the further device.

5. The method according to claim 1, wherein: The information obtained from the further device further indicates an object class of the object.

6. The method according to claim 5, wherein: The information obtained from the further device further indicates a confidence level associated with the object category.

7. The method according to claim 1, wherein: In the act of lowering, the threshold is lowered in a region along the periphery of the field of view.

8. The method according to claim 1, wherein: The further device is one of the following: Another camera; IR camera; Thermal cameras; radar; Sonar; and LiDAR.

9. The method according to claim 1, further comprising: After the object is detected in the field of view of the camera, increasing the threshold; as well as In case the object is detected using the increased threshold: maintaining a mask of the detected object in the video stream; In case the object is not detected using the increased threshold: Masking of the detected object in the video stream is discontinued.

10. The method according to claim 9, wherein: The action of increasing the threshold is performed at a predetermined time after the object is detected.

11. A non-transitory computer-readable storage medium having instructions stored thereon for implementing the method of claim 1 when executed in a camera having a processor.

12. A camera for masking a detected object in a video stream captured by the camera, the camera comprising: Circuitry configured to perform the following functions when the camera is arranged in a camera system comprising the camera and at least a further device, wherein for the further device a position and a field of view are known, and for the camera a position and a field of view are known, and wherein the field of view of the further device and the field of view of the camera are non-overlapping: an obtaining function configured to obtain information indicating that an object is approaching the field of view of the camera, wherein the obtained information is determined based on information indicating the position and movement direction of the object from the further device and the known positions and fields of view of the camera and the further device; a lowering function configured to lower a threshold for detecting an object to be masked in the video stream captured by the camera in response to the information obtained, wherein the threshold specifies a confidence level above which an object is determined to belong to a category of objects to be masked in the video stream, or wherein the threshold is a threshold for what is detected as a foreground and what is detected as a background, and specifies an amount of change in image data in an image frame relative to image data in a previous image frame, above which an object is determined to belong to the foreground; a detection function configured to detect an object to be masked in the video stream using the reduced threshold; and An insertion function is configured to insert a detected mask of the object in the video stream.

Images