Method and system for providing video compression stream

Abstract

The present disclosure relates to a method for providing a video compression stream. The background blurring system identifies at least a first pixel in the background at a first time point; determines a blurred pixel value of at least the first pixel; provides the blurred pixel value in the video compression stream; identifies at least the first pixel as having become in the foreground at a second time point thereafter; provides the identified non-blurred pixel value of at least the first pixel in the video compression stream after the second time point; identifies at least the first pixel as having become included in the background at a third time point thereafter, wherein the value of at least the first pixel remains unchanged from the intermediate time point to the third time point; and provides the blurred pixel value in the video compression stream after the third time point when the third time point is within a background merging time period from the intermediate time point. The disclosure also relates to such a background blurring system, a camera including such a system, a corresponding computer program product, and a non-volatile computer-readable storage medium.

Description

Technical Field

[0001] The present disclosure is directed to providing a video compression stream from a camera adapted to capture a scene. Background Art

[0002] By using a camera (e.g., for surveillance purposes), it is possible to view what is happening in a particular environment (e.g., an office, an industrial site, a part of a city, an open square, a city block, a road, etc.) while it is happening and / or at a later time. A video stream of the surrounding environment (i.e., a scene) can help provide a view and information related to the scene, which can be revealing and / or interesting in one way or another. However, providing a video stream such as a real-time video stream from a camera generally results in a high computational cost and further requires a large amount of bit rate. Therefore, it is known to use various known techniques to compress such a video stream. It is also known that they can be provided unclearly by blurring their background portions, so that one or more people and / or objects in the foreground can be kept in focus, while the background can be rendered out of focus. However, although background blurring (in addition to possible efficiency in terms of privacy) may result in reduced bit rate consumption, reduced computational costs and / or bit rates in the video stream, this remains a challenge. Summary of the invention

[0003] It is therefore an object of the present invention to provide a method for providing a video compression stream, for example coming from a camera adapted to capture a scene, in an improved and / or alternative manner.

[0004] The above objects can be achieved by the subject matter disclosed herein. Embodiments are set forth in the accompanying claims, the following description and the accompanying drawings.

[0005] The disclosed subject matter relates to a method for providing a video compression stream from a camera adapted to capture a scene, performed by a background blur system. The background blur system identifies, at a first time point, in a first image of the scene, at least a first pixel included in a background of the first image. The background blur system further determines a blurred pixel value for the at least first pixel. In addition, the background blur system provides the blurred pixel value in the video compression stream after the first time point. The background blur system also identifies, at a second time point thereafter, in a second image of the scene, that at least the first pixel has changed to be included in a foreground of the second image. In addition, the background blur system continuously and / or intermittently provides the identified non-blurred pixel value for the at least first pixel in the video compression stream after the second time point. The background blur system further identifies, at a third time point thereafter, in a third image of the scene, that at least the first pixel has changed to be included in a background of the third image, wherein the value of the at least first pixel remains unchanged from the intermediate time point to the third time point. In addition, when the third time point is within a predeterminable background merging time period from the intermediate time point, the background blur system provides the blurred pixel value in the video compression stream after the third time point.

[0006] The disclosed subject matter further relates to a background blur system for providing a video compression stream from a camera adapted to capture a scene. The background blur system comprises a first identification unit for and / or adapted to identify, in a first image of the scene, at least a first pixel included in the background of the first image at a first point in time. The background blur system further comprises a first blur unit for and / or adapted to determine a blurred pixel value for at least the first pixel. In addition, the background blur system comprises a first providing unit for and / or adapted to provide the blurred pixel value in the video compression stream after the first point in time. The background blur system further comprises a second identification unit for and / or adapted to identify, in a second image of the scene at a second point in time thereafter, that at least the first pixel has changed to be included in the foreground of the second image. In addition, the background blur system comprises a second providing unit for and / or adapted to continuously and / or intermittently provide the identified non-blurred pixel value for at least the first pixel in the video compression stream after the second point in time. The background blurring system further comprises a third identification unit for and / or adapted to identify, at a later third time point, in a third image of the scene that at least a first pixel has been changed to be included in the background of the third image, wherein the value of at least the first pixel remains unchanged from the intermediate time point to the third time point. In addition, the background blurring system comprises a third providing unit for and / or adapted to provide the blurred pixel value in the video compression stream after the third time point when the third time point is within a predeterminable background merging time period from the intermediate time point.

[0007] Furthermore, the disclosed subject matter relates to a camera comprising a background blur system as described herein.

[0008] Furthermore, the disclosed subject matter relates to a computer program product comprising a computer program stored on a computer readable medium or carrier wave, the computer program containing computer program code means arranged to cause a computer or a processor to perform the steps of the background blurring system described herein.

[0009] The disclosed subject matter also relates to a non-volatile computer-readable storage medium having stored thereon the computer program product.

[0010] Thus, a method is introduced, according to which computationally efficient and / or bitrate efficient background blurring can be achieved. That is, due to identifying at a first point in time in a first image of a scene captured by a camera at least a first pixel included in the background of the first image, it is determined that one or more pixels are considered to belong to the background area of ​​the captured scene image at said point in time, for example, due to the scene including static and / or motionless people and / or objects. Furthermore, that is, due to determining a blurred pixel value for at least the first pixel, a blurred value for at least the first identified background pixel can be determined, which blurred value can then cause the pixel to be perceived as blurred, unfocused and / or unsharp. Moreover, that is, due to providing the blurred pixel value in the video compression stream after the first point in time, the determined blurred pixel value of at least the first pixel associated with the first captured image is used in the video compression stream. In this way, by using blurred pixel values ​​instead of non-blurred pixel values ​​for one or more identified background pixels in the video compression stream, bitrate can be reduced and / or privacy requirements can be met. Furthermore, since the background pixels are blurred as opposed to the non-blurred foreground pixels, objects including foreground pixels (such as, for example, moving people and / or objects) are in focus and / or appear emphasized compared to non-moving people and / or objects including background pixels. Furthermore, that is, since at a later second point in time, at least a first pixel is identified in a second image of the scene as having changed to be included in the foreground of the second image, it is determined that in the second image of the scene appearing at any second point in time, one or more pixels previously identified as belonging to the background are considered to have changed to belonging to the foreground area of ​​the captured second image, for example, because one or more people and / or objects entered the scene. Furthermore, that is, since after the second point in time, the identified non-blurred pixel value for at least the first pixel is continuously and / or intermittently provided in the video compression stream, the identified pixel value of at least the first pixel associated with the second captured image and its other identified changed pixel values ​​in any number of later captured images of the scene are used in the video compression stream. Thus, by using continuously and / or intermittently updated non-blurred pixel values ​​for one or more identified foreground pixels in a video compression stream, foreground pixels (such as, for example, one or more moving people and / or objects) are brought into focus compared to blurred background pixels.In addition, that is, since at a later third time point, at least a first pixel is identified in a third image of the scene as having changed to be included in the background of the third image, wherein the value of at least the first pixel remains unchanged from the intermediate time point to the third time point, it is determined that one or more pixels in the third image of the scene occurring at any third time point, which were previously identified as belonging to the background and then identified as belonging to the foreground, are changed again to belong to the background area (after remaining unchanged since the intermediate time point), for example, this is caused by one or more people and / or objects leaving the scene and / or at least partially standing still. In addition, that is, since after the third time point (when the third time point is within a predeterminable background merging time period from the intermediate time point), a blurred pixel value is provided in the video compression stream, so after the third time point, if the time range from the intermediate time point (i.e., when at least the first pixel last changed its pixel value) to the third time point (i.e., when at least the first pixel changes from a foreground pixel to a background pixel again) is within a set time period, the previously determined blurred pixel value for at least the first pixel is used again in the video compression stream. That is, if within a predeterminable time period, at least the first pixel changes from an unchanged foreground pixel to a background pixel again (for example, this can be exemplified by one or more people and / or objects passing through the scene within the background merging time period (for example, set to an exemplary five seconds) and / or standing still for no longer than the background merging time period), the already determined blur value (determined after the first point in time (i.e., when at least the first pixel was previously identified as a background pixel)) is used again in the video compression stream. Therefore, by reusing the determined blur pixel value (which is considered to be still valid when the background merging time period criterion is met), it can be considered unnecessary to determine an updated blur pixel value for at least the first pixel, thereby providing reduced computational cost and / or efficient bit rate compression of the video compression stream.

[0011] To this end, a method for providing a video compression stream (eg, a video compression stream from a camera adapted to capture a scene) in an improved and / or alternative manner is introduced.

[0012] The technical features and corresponding advantages of the above method will be discussed in further detail below. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Various aspects of non-limiting embodiments including particular features and advantages will be readily understood from the following detailed description and accompanying drawings, in which:

[0014] Figure 1 a to Figure 1 e is a schematic diagram showing an exemplary video compression stream image provided from an exemplary background blur system according to an embodiment of the present disclosure;

[0015] Figure 2 is a schematic block diagram illustrating an exemplary background blur system according to examples of the present disclosure; and

[0016] Figure 3 is a flow chart depicting an exemplary method performed by a background blur system according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0017] The non-limiting embodiments of the present disclosure will now be described more fully below with reference to the accompanying drawings, in which the currently preferred embodiments of the present disclosure are shown. However, the present disclosure can be implemented in many different forms and should not be construed as being limited to the embodiments set forth herein. Throughout the text, similar reference numerals indicate similar elements. The dotted lines of certain boxes in the figure indicate that these units or actions are optional, rather than mandatory.

[0018] In the following, according to embodiments herein involving providing a video compression stream from a camera adapted to capture a scene, a method will be disclosed according to which computationally efficient and / or bitrate efficient background blurring can be achieved.

[0019] Now refer to the accompanying drawings, specifically Figure 1 to Figure 2 , Figure 1 a to Figure 1 e depicts a schematic diagram of an exemplary video compression stream image provided from an exemplary background blurring system 1 according to an embodiment of the present disclosure, and Figure 2 A schematic block diagram illustrating an exemplary background blurring system 1 according to an embodiment of the present disclosure is depicted. The background blurring system 1 is adapted and / or configured to provide a video compression stream from a camera 2 adapted to capture a scene.

[0020] Camera 2 may refer to any arbitrary (e.g., known) camera suitable for image capture. According to an example, the camera 2 may further be included in a camera arrangement and / or network (not shown) including additional components, such as part of an exemplary surveillance system (not shown). The scene that the camera 2 is suitable for capturing may be represented by any arbitrary scene, such as an office, an industrial site, a part of a city, an open square, a city block, a road, etc. The scene may further optionally be a scene of interest to be monitored. In addition, the camera 2 may be arranged relatively close to one or more potential, such as people and / or objects in the scene, so that the objects and / or people constitute a relatively large part of the captured image of the scene. The background blur system 1 may be included in the camera 2; or, the background blur system 1 may, for example, be at least partially arranged outside the camera 2, such as in an exemplary camera arrangement or surveillance system, and / or in one or more servers. In addition, the camera 2 and / or the background blur system 1 may include components that support pixel blurring. Considering the limitations of the features provided herein, the video compression stream to be provided may have any feasible size and / or length, and may also have any feasible format. The phrase "background blur system" may refer to a "processing system", a "compression system" and / or a "video stream background blur system", and "for providing a video compression stream" may refer to "for providing, producing, generating, displaying, storing and / or saving a video compression stream", "for background blur processing of a video stream" and / or "for processing of a video stream". On the other hand, throughout the disclosure, the phrase "video compression stream" may refer to a "compressed video stream" and / or a "compressed video signal", and according to an example, further indicates a "video compression real-time stream". In addition, a "camera suitable for capturing a scene" may refer to a "camera suitable for capturing the surrounding environment", and according to an example, further refers to a "surveillance camera suitable for capturing a scene" and / or a "camera suitable for capturing a scene to be monitored".

[0021] The background blur system 1 is suitable and / or configured to identify, at a first point in time, in a first image of a scene, at least a first pixel included in the background of the first image, for example by means of a first recognition unit 101. Thereby, at said point in time, one or more pixels of the background area of ​​the captured image that are considered to belong to the scene are determined. According to the example, one or more pixels considered to be background pixels may, for example, originate from a scene including static and / or motionless people and / or objects. It should be pointed out that these people and / or objects, for example, constituting background pixels, are not necessarily farther away from the camera 2 than the people and / or objects constituting foreground pixels, they may also be closer to the camera 2 than the people and / or objects constituting foreground pixels.

[0022] The first image may refer to any arbitrary image of a scene captured by the camera 2, and in a similar manner, the first point in time may refer to any arbitrary point in time. Similarly, at least the first pixel may refer to any arbitrary one or more pixels in the first image located at any arbitrary position in the image. In addition, the conclusion that at least the first pixel is included in the background of the first image may be determined and / or has been determined by any arbitrary, for example known, feasible manner, such as by detecting, for example, pixel values ​​and / or pixel intensities that are constant within a predeterminable range, and / or by generally known image depth detection, for example with the support of IR imaging and / or depth cameras of the scene. That is, pixels of an image may be considered as foreground pixels or background pixels, wherein, according to an example, foreground pixels may indicate pixels that are dynamic and / or changing (and / or have dynamic and / or changing values), while background pixels may indicate pixels that remain static and / or unchanged (and / or have values ​​that remain static and / or unchanged within a predeterminable range), for example. The term “identified in an image” may refer throughout the present disclosure to “detected, filtered out and / or determined in an image”, while “in an […] image” may refer throughout the present disclosure to “in an […] captured image” and / or “in an […] image captured by the camera”. On the other hand, the phrase “at least a first pixel” may refer to “one or more pixels” and / or “one or more adjacent and / or distributed pixels”, while “pixels included in the background” may refer to “pixels regarded and / or determined as included in the background”, “pixels included in the determined background”, “pixels belonging to the background” and / or “pixels included in the background area”. According to an example, the phrase “pixels included in the background of an […] image” may refer throughout the present disclosure to “pixels included in a static and / or unchanging area of ​​an […] image” and / or “pixels determined and / or regarded as being static and / or unchanging within a predeterminable range (and / or having a value that is static and / or unchanging within a predeterminable range) and / or pixels that remain static and / or unchanging (and / or have a value that remains static and / or unchanging)”.

[0023] The background blurring system 1 is adapted and / or configured to determine a blurred pixel value for at least a first pixel, e.g. by means of a first blurring unit 102. Thus, a blur value may be determined for at least a first identified background pixel, which blur value then causes said pixel to be perceived as blurred, unfocused and / or unsharp.

[0024] Determining the blurred pixel value for one or more identified background pixels can be done in any arbitrarily known, for example, feasible manner, for example with the aid of one or more generally known image filter kernels (e.g., blur kernels). Such exemplary image filter kernels can have any arbitrarily feasible dimensions and / or sizes, for example, referring to square kernels, such as 3*3 kernels, 11*11 kernels, 21*21 kernels, etc., and / or indicating non-square kernels, such as 21*11 kernels, where 21 can be set horizontally. The degree of blurring of at least the first pixel can be arbitrarily selected, for example, as deemed suitable for the upcoming implementation. Optionally, for bit rate efficiency, the kernel size of the image filter kernel used to determine the blurred pixel value of at least the first pixel can be consistent with the macroblock size of the I frame of the video compression stream. The phrase "determining" a blurred pixel value can refer to "applying" a blurred pixel value and / or "calculating" a blurred pixel value throughout the present disclosure, and according to the example, also refers to "determining" a blurred pixel value based on an image filter kernel. On the other hand, the phrase "blurred pixel value" may refer throughout the present disclosure to "unfocused and / or unsharp pixel value", while "for at least the first pixel" may indicate "applicable to at least the first pixel". According to an example, "blurred pixel value" may additionally or alternatively refer to "pixel value with color removed". Furthermore, according to an example, "blurred" may refer throughout the present disclosure to "blurred and / or modified" or "only modified", whereby the "modified" pixel value may then be modified in any arbitrary feasible manner.

[0025] The background blurring system 1 is adapted and / or configured to provide the blurred pixel value in the video compression stream after the first point in time, e.g. by means of the first providing unit 103. Thereby, the determined blurred pixel value for at least the first pixel related to the first captured image is used in the video compression stream, e.g. Figure 1 a, for displaying corresponding at least first stream pixels 111 in a corresponding first compressed image 11 of the stream. Thus, by using blurred pixel values ​​rather than non-blurred pixel values ​​for one or more identified background pixels in a video compression stream, bit rate can be reduced and / or privacy requirements can be met. Furthermore, because background pixels are blurred in contrast to non-blurred foreground pixels, objects including foreground pixels (such as, for example, moving people and / or objects) can be focused and / or highlighted compared to non-moving people and / or objects including background pixels. Moving people and / or objects including background pixels. In the exemplary Figure 1 In a, all pixels of the first image (including at least the first pixel) are considered to belong to the background, so in the first compressed image 11, all pixels are blurred.

[0026] The blurred pixel values ​​may be provided in the video compression stream in any arbitrary, e.g. known, practicable manner, e.g. displayed on, e.g. one or more displays 3 communicating directly or indirectly with the background blur system 1 and / or the video camera 2, and / or recorded and / or stored, e.g. digitally, for later use. The phrase "providing […] pixel values ​​in the video compression stream" may refer throughout the present disclosure to "inserting, setting, utilizing and / or using […] pixel values ​​in the video compression stream" and / or "storing, saving and / or recording […] pixel values ​​for use in the video compression stream", and according to examples, also to "continuously and / or intermittently providing […] pixel values ​​in the video compression stream" and / or "displaying […] pixel values" and / or "utilizing […] pixel values ​​in one or more subsequent frames of the video compression stream". On the other hand, the phrase "providing after the […] point in time" may refer throughout the present disclosure to "providing after the […] point in time and / or essentially at the point in time".

[0027] The background blurring system 1 is adapted and / or configured to recognize, at a later second time point, in the second image of the scene that at least the first pixel has been changed to be included in the foreground of the second image, for example by means of the second recognition unit 104. Thus, it can be determined that in the second image of the scene appearing at any second time point, one or more pixels previously recognized as belonging to the background are considered to have been changed to belong to the foreground area of ​​the captured second image. According to an example, this may be caused, for example, by one or more persons and / or objects entering the scene.

[0028] The second image may refer to any arbitrary image of a scene captured by the camera 2 after the previously captured first image, and in a similar manner, the second point in time may refer to any arbitrary point in time occurring after the first point in time, for example from a few milliseconds to a few minutes or even hours thereafter. It should be noted that the second image is not necessarily an image directly after the first image. On the contrary, between the so-called first and second images, any number of consecutive images may be captured. Furthermore, the conclusion that at least the first pixel is included in the foreground of the second image may be determined and / or has been determined in any arbitrary, for example known, feasible manner, for example by detecting changing or substantially changing pixel values ​​and / or pixel intensities, and / or as previously indicated by generally known image depth detection. The phrase "the pixel has been changed to be included in the foreground" may mean "the pixel has been changed to be included in the foreground" and / or "the pixel has been determined and / or regarded as having been changed to be included in the foreground", and "the pixel has been changed to be included in the foreground" may mean "the pixel has been changed to be included in the determined foreground", "the pixel has been changed to belong to the foreground" and / or "the pixel has been changed to be included in the foreground area". According to an example, the phrase "the pixel has been changed to be included in the foreground of the second image" may mean "the pixel has been changed to be included in a dynamic, changing and / or substantially changing area of ​​the second image", "the pixel is determined and / or regarded as having changed to dynamic, changing and / or substantially changing (and / or having a value that has changed to dynamic, changing and / or substantially changing)", and / or "the pixel is determined and / or regarded as currently or substantially currently having changed to dynamic, changing and / or substantially changing (and / or having a value that has changed to dynamic, changing and / or substantially changing)". In addition, the term "second image" may indicate "a second image thereafter" and / or "a second image after the first image".

[0029] The background blurring system 1 is adapted and / or configured to provide, after a second point in time, continuously and / or intermittently in the video compression stream the identified non-blurred pixel value of at least the first pixel, e.g. by means of the second providing unit 105. Thereby, for the video compression stream, the identified pixel value of at least the first pixel related to the second captured image and its other identified changed pixel values ​​in any number of subsequent captured images of the scene are used, e.g. Figure 1 b, for displaying corresponding at least first stream pixels 111 in corresponding second compressed images 12 of the stream. Thus, by utilizing the continuously and / or intermittently updated non-blurred pixel values ​​of one or more identified foreground pixels in the video compression stream, foreground pixels such as one or more moving persons and / or objects can be brought into focus compared to blurred background pixels. Figure 1b, at least the first pixel of the second image is considered to belong to the foreground, and therefore, in the second compressed image 12, the pixel (here belonging to the exemplary moving person) is unblurred, i.e., sharp. It should be noted that in the example shown, since the person continues to move, at least the first pixel may change continuously and / or intermittently, whereby the unblurred pixel value of the at least first pixel may change continuously and / or intermittently, for example, as the images of the scene are captured continuously.

[0030] The non-blurred pixel value of at least the first pixel may be identified and / or derived in any arbitrarily known, e.g. feasible manner, e.g. with the support of generally known image processing. The phrase "providing [...] an identified non-blurred pixel value" may mean "providing [...] a captured and / or derived non-blurred pixel value", "providing [...] an updated non-blurred pixel value" and / or "providing [...] an updated identified non-blurred pixel value". Furthermore, according to the example, the phrase "providing [...] a continuously and / or intermittently identified non-blurred pixel value of the at least first pixel" may mean "providing [...] an identified non-blurred pixel value of the at least first pixel as long as the at least first pixel is included in the foreground of the continuously captured images. On the other hand, the phrase "non-blurred pixel value" may mean "an actual, detected, focused and / or sharp pixel value".

[0031] The background blurring system 1 is adapted and / or configured, for example by means of the third recognition unit 106, to recognize, at a later third point in time, in a third image of the scene, that at least a first pixel has been changed to be included in the background of the third image, wherein from the intermediate point in time to the third point in time, at least the value of the first pixel remains unchanged. It is thereby determined that one or more pixels in the third image of the scene appearing at any third point in time, which were previously identified as belonging to the background and then identified as belonging to the foreground, are considered to be changed again (after remaining unchanged since the intermediate point in time) to belong to the background area. According to the example, this is caused, for example, by one or more persons and / or objects leaving the scene and / or being at least partially stationary.

[0032] The third image may refer to any arbitrary image of the scene captured by the camera 2 after the previously captured second image, and in a similar manner, the third time point may indicate any arbitrary time point occurring after the second time point, for example a few milliseconds to minutes or even hours thereafter. It should be noted that the third image is not necessarily an image directly after the second image. Instead, between the so-called second and third images, any number of consecutive images may be captured, for example in the following case: namely, at least the first pixel remains in the foreground for a period of time, for example, with or without a changing pixel value such as caused by an exemplary moving person. The conclusion that at least the first pixel is included in the background of the third image after the third time point may be determined and / or has been determined in any arbitrary, for example known, feasible manner, for example by detecting, deducing and / or knowing that the pixel value of at least the first pixel remains unchanged and / or substantially unchanged within a predeterminable range, for example just before the third time point. Here, at least the first pixel is considered to remain unchanged since an exemplary intermediate time point, which may occur at any arbitrary time point after the second time point, for example a few milliseconds to minutes or even hours thereafter.

[0033] The phrase “the pixel has been changed to be included in the background” may mean “the pixel has been changed to be included in the background again”, “the pixel has been changed to be included in the background” and / or “the pixel has been determined and / or regarded as having been changed to be included in the background”, and further means “the pixel has been changed to be included in the determined background”, “the pixel has been changed to belong to the background” and / or “the pixel has been changed to be included in the background area”. According to an example, the phrase “the pixel has been changed to be included in the background of the third image” may mean “the pixel has been changed to be included in the static and / or unchanged area of ​​the third image” and / or “the pixel is determined and / or regarded as having changed to static and / or unchanged within a predeterminable range, and / or remains static and / or unchanged (and / or has a value that has changed to static and / or unchanged within a predeterminable range, and / or remains static and / or unchanged)”. In addition, the term “third image” may indicate “a third image thereafter” and / or “a third image after the second image”. Furthermore, the expression "at least the value of the first pixel" may mean "at least the pixel value of the first pixel", and "remains unchanged" may mean "remains essentially unchanged", "has reflected no movement and / or no motion" and / or "remains unchanged within a predeterminable range". On the other hand, the expression "remains unchanged from the intermediate time point to the third time point" may mean "remains unchanged from the intermediate time point at least until the third time point" and / or "remains unchanged from the intermediate time point after the second time point until the third time point".

[0034] The background blurring system 1 is adapted and / or configured to provide the blurred pixel value in the video compression stream after a third time point (when the third point is within a predeterminable background merging time period from the intermediate time point), e.g. by means of the third providing unit 107. Thus, after the third time point, if the time range from the intermediate time point (i.e., when at least the first pixel last changed its pixel value) to the third time point (i.e., when at least the first pixel changes from a foreground pixel to a background pixel again) is within the set time period, the previously determined blurred pixel value of at least the first pixel is used again in the video compression stream, e.g. Figure 1 c, for displaying the corresponding at least first stream pixel 111 in the corresponding third compressed image 13 of the stream. That is, if, within a predeterminable time period, at least the first pixel changes from a constant foreground pixel to a background pixel again (for example, it can be exemplified that one or more people and / or objects pass through the scene within the background merging time period (for example, set to an exemplary five seconds) and / or stand still for no longer than the background merging time period), the already determined blur value (which is determined after the first time point (i.e., when at least the first pixel was previously identified as a background pixel)) is reused in the video compression stream. Therefore, by reusing the determined blur pixel value (which is considered to be still valid when the background merging time period criterion is met), it can be considered unnecessary to determine an updated blur pixel value for at least the first pixel, thereby providing reduced computational cost and / or efficient bit rate compression of the video compression stream. In addition, since the amount of computation required thereafter is reduced, a relatively large image filter kernel (for example, represented by a blur kernel) can be used. In the exemplary Figure 1 c, at least the first pixel of the third image is considered to belong to the background again and, since this change (according to this example) occurs within the set background merging time period, in the third compressed image 13, said pixel is blurred using the previously determined blur value.

[0035] The background merging time period may be any arbitrarily feasible time range, such as a fixed time range, such as less than one second to tens of seconds or even to several minutes or hours. According to one example, the background merging time period is greater than two seconds, according to another example, greater than five seconds, and according to yet another example, greater than ten seconds. Similarly, according to an example, the background merging time period is less than sixty seconds, according to another example, less than thirty seconds, and according to yet another example, less than ten seconds. The background merging time period may be further varied and / or adjusted, such as being set to be considered suitable for upcoming implementations and / or situations. The phrase "when the third time point is within a predeterminable background merging time period from the intermediate time point" may indicate "if, if / if the third time point is within a predeterminable background merging time period from the intermediate time period", and "predeterminable background merging time period" may indicate "predetermined background merging time period" and / or simply "background merging time period". In addition, the phrase "background merging time period" may indicate "first time period", "reuse threshold time period" and / or "background merging time range". According to an example, the phrase “after the third time point, providing the blurred pixel value in the video compression stream” may mean “after the third time point, reusing the blurred pixel value in one or more subsequent frames of the video compression stream”.

[0036] Optionally, the background blurring system 1 is adapted and / or configured, for example by means of the optional second blurring unit 108, to determine an updated blurred pixel value for at least the first pixel when the third time point is not within the background merging time period. Thus, if the time range from the intermediate time point (i.e., from this time on, the value of at least the first pixel remains unchanged) to the third time point (i.e., by this time, at least the first pixel changes from a foreground pixel to a background pixel again) is not within the set time period, an updated blurred value may be determined for at least the first pixel. That is, if the background merging time period criterion is not met, determining an updated blurred pixel value for at least the first pixel is deemed to be necessary, because the previously determined blurred pixel value is deemed to be no longer relevant, since at least the first pixel is deemed to reflect a different, new and / or updated background compared to the previously determined blurred pixel value. The phrase "when the third time point is not within the background merging time period" may mean "if, were / assuming that the third time point is not within the background merging time period", and "updated blurred pixel value" may mean "new blurred pixel value".

[0037] Further optionally, the background blurring system 1 is then adapted and / or configured to provide an updated blurred pixel value in the video compression stream after a third point in time, for example by means of an optional fourth providing unit 109. Thus, after said third point in time, if the time range from the intermediate point in time (i.e., from this point onwards, the value of at least the first pixel remains unchanged) to the third point in time (i.e., when at least the first pixel changes from a foreground pixel to a background pixel again) is not within the set time period as described above, the updated determined blurred pixel value for at least the first pixel is used in the video compression stream, for example as Figure 1 e, for displaying the corresponding at least first stream pixel 111 in the corresponding alternative third compressed image 13' of the stream. That is, if the time range from when at least the first pixel becomes an unchanged foreground pixel to then becomes a background pixel again is not within the set time period (for example, it can be exemplified as follows, an object such as a person enters the scene and then stands at least partially still, such as Figure 1 d), then the determined updated blurred pixel value is used in the video compression stream instead of the previously determined blurred pixel value (which is now outdated and / or considered outdated). Figure 1 e, at least the first pixel is considered to belong to the background again and since this change (according to the example described above) is shown by, for example, a person entering the scene and then standing still at least partially for longer than the set background merging time period, in the alternative third compressed image 13' said pixel is blurred again and now has an updated determined blurred pixel value.

[0038] In addition, optionally, the background blurring system 1 is further adapted and / or configured to additionally determine updated blurred pixel values ​​for additional pixels in the vicinity of at least the first pixel, e.g. by means of an optional second blurring unit 108. The background blurring system 1 is then further adapted and / or configured to additionally provide updated blurred pixel values ​​for additional pixels, e.g. by means of an optional fourth providing unit 109. Thus, after the third point in time discussed previously, in addition to determining and subsequently providing the updated blurred pixel value for at least the first pixel in the video compression stream, updated pixel values ​​are also determined and subsequently provided for a predeterminable portion of pixels in the vicinity of at least the first pixel and / or associated with at least the first pixel.

[0039] Additional pixels near at least the first pixel for which the updated blurred pixel value may be additionally determined may be represented by any feasible arbitrary dimensional portion and / or number of pixels near at least the first pixel (e.g., pixels adjacent to the at least first pixel). According to an example, the additional pixels may be represented by pixels forming a rectangle or quadrat (e.g., 10*10 pixels) around at least the first pixel. It should be noted that at least the first pixel does not necessarily have to be located at the center of the rectangle or quadrat. According to another example, the additional pixels may be represented by pixels identified as a part of the object to which at least the first pixel belongs and / or associated with the object. The phrase "for additional pixels near the at least first pixel" may refer to "for additional pixels surrounding and / or adjacent to the at least first pixel" and / or "for a portion of additional pixels near the at least first pixel".

[0040] Optionally, the background blurring system 1 is adapted and / or configured, for example by means of an optional supplementary blurring unit 110, to determine an updated blurred pixel value of at least a first pixel from the current image of the scene when the blurred pixel value at a predetermined supplementary time point is determined to have not been updated within a predetermined supplementary interval. Then, the background blurring system 1 is further adapted and / or configured, for example by means of an optional supplementary providing unit 111, to provide the updated blurred pixel value in the video compression stream after the supplementary time point. Therefore, after the supplementary time point, if the blurred pixel value at any time point is determined not to have been updated within a set time period, the updated blurred pixel value of at least the first pixel is determined from the current image and then used in the video compression stream. Therefore, provided that at least the first pixel remains a background pixel within the set supplementary time period, the updated blurred pixel value can still be determined and provided in the video compression stream. Therefore, situations involving, for example, providing previously determined blurred pixel values ​​(which may have been outdated and / or erroneous) can be compensated. In a similar manner, situations involving, for example, slowly changing light conditions (eg, light conditions at dusk or dawn, which do not necessarily result in at least the first pixel changing from background to foreground) may additionally or alternatively be compensated.

[0041] The supplementary time period may be any arbitrarily feasible time range, for example, from less than one second to tens of seconds or even to several minutes or hours. According to an example, the supplementary time period may be represented by a time range equal to or substantially equal to the time range of the background merging time period discussed previously. The supplementary time period may be further varied and / or adjusted, for example, set to be suitable for the upcoming implementation and / or situation. In a similar manner, the supplementary time point may be any arbitrarily feasible time point. According to an example, the supplementary time point may be represented by a time point that occurs after the supplementary time point discussed previously and through a time range equal to or substantially equal to the background merging time period. It should be noted that the supplementary time period and / or its size may be different for different pixels. The phrase "when the blurred pixel value at a predeterminable supplementary time point" may refer to "assuming and / or assuming the blurred pixel value at a predeterminable supplementary time point". On the other hand, the phrase "predeterminable supplementary time point" may refer to "predetermined supplementary time point" and / or only indicate "supplementary time point", while "predeterminable supplementary time period" may indicate "predetermined supplementary time period" and / or only indicate "supplementary time period". Furthermore, the phrase "from a current image of the scene" may mean "from a substantially current and / or upcoming image of the scene".

[0042] Optionally, the background blurring system 1 is adapted and / or configured, e.g. by means of an optional I-frame blurring unit 112, to determine updated blurred pixel values ​​for at least the first pixel and other pixels identified as a predetermined portion of pixels included in the background of the current image of the scene when the predetermined upcoming frame of the video compression stream at a predetermined I-frame-related time point is an I-frame. Then, the background blurring system 1 can be further adapted and / or configured, e.g. by means of an optional I-frame providing unit 113, to provide the updated blurred pixel values ​​in the video compression stream after the I-frame-related time point. Thus, after the I-frame-related time point, if the predetermined upcoming frame of the video compression stream is an I-frame, updated blurred pixel values ​​can be determined from the current image for at least the first pixel and additionally for a portion of pixels selected from the background pixels and then used in the video compression stream. Thus, for one or more selected I-frames, a series of selected background pixels including at least the first pixel can be updated. Thus, updated blurred pixel values ​​for selected pixels (which pixel selection may differ from one selected I-frame to another selected I-frame) may be determined and provided without negatively impacting computational cost and / or video compression stream bitrate.

[0043] The predeterminable I-frame-related time point may be represented by any arbitrarily feasible time point, and similarly, the predeterminable upcoming I-frame may refer to any arbitrarily feasible I-frame of the video compression stream. For example, every n I-frame may be a selected (i.e., predeterminable) -I-frame, where n may be arbitrarily selected as deemed appropriate for the upcoming implementation and / or situation. The phrase "when […] the predeterminable upcoming frame of the video compression stream is an I-frame" may refer to "if, were / assuming that […] the predeterminable upcoming frame of the video compression stream is an I-frame", and "the upcoming frame" may indicate "the upcoming frame" and / or "the next upcoming frame". On the other hand, the phrase "the predeterminable upcoming frame" may refer to "the predetermined upcoming frame" and / or "the selected upcoming frame", and on the other hand, "the predeterminable I-frame-related time point" may refer to "the predetermined I-frame-related time point" and / or simply "the I-frame-related time point". Furthermore, the phrase “other pixels of the predetermined portion” may refer to “additional pixels of the predetermined portion”, “other pixels of the predetermined portion” and / or simply “a portion of other pixels”, and the phrase “a portion identified as being out of the pixels included in the background” may refer to “a range, a set and / or a selection of pixels identified as being out of the pixels included in the background”. Furthermore, the phrase “of the current image of the scene” may refer to “of the substantially current and / or upcoming image of the scene”.

[0044] like Figure 2As further shown, the background blur system 1 may include a first recognition unit 101, a first blurring unit 102, a first providing unit 103, a second recognition unit 104, a second providing unit 105, a third recognition unit 106, a third providing unit 107, an optional second blurring unit 108, an optional fourth providing unit 109, an optional supplementary blurring unit 110, an optional supplementary providing unit 111, an optional I-frame blurring unit 112, and an optional I-frame providing unit 113, which have been described in detail above. In addition, the embodiments of the present invention for providing a video compression stream from a camera 2 suitable for capturing a scene can be implemented by one or more processors (e.g., a processor 114 (e.g., a graphics processing unit GPU and / or a central processing unit CPU)) together with computer program codes for performing the functions and actions of the embodiments of the present invention. The program code can also be provided as a computer program product, for example in the form of a data carrier carrying a computer program code, which is used to perform the embodiments of the present invention when loaded into the background blur system 1. One such carrier may be in the form of a CD ROM disk, an optical disk, a solid state disk, a flash memory and / or a hard drive, but other data carriers are also feasible. In addition, the computer program code may be provided as pure program code on a server and downloaded to the background blurring system 1. The background blurring system 1 may also include a memory 115, which includes one or more storage units. The memory 115 may be arranged to store, for example, information, and further store data, configurations, schedules and applications so that when these information, data, configurations, schedules and applications are run in the background blurring system 1, the method of the present invention can be performed. For example, the computer program code may be implemented in firmware stored in the flash memory 115 of the embedded processor 114. In addition, the units 101-113, the optional processor 114 and / or the optional memory 115 may be at least partially included in the camera 2 and / or in a surveillance system that, for example, optionally includes the camera 2, is associated with them, and / or is connected to them. Those skilled in the art will also understand that the units 101-113 described above may refer to a combination of analog and digital circuits and / or one or more processors configured by, for example, software and / or firmware stored in a memory (such as memory 115), which, when executed by one or more processors (such as processor 114), performs as described. One or more of these processors and other digital hardware may be included in a single application-specific integrated circuit ASIC, or multiple processors and various digital hardware may be distributed among several separate components, whether these components are packaged separately or assembled into a system-on-chip Soc.

[0045] Figure 3is a flow chart showing an exemplary method performed by a background blur system 1 according to an embodiment of the present disclosure. The method is for providing a video compression stream from a camera 2 adapted to capture a scene. The exemplary method, which may be repeated continuously, comprises: Figure 1 to Figure 2 In addition, where applicable, the actions may be performed in any suitable order and / or one or more actions may be performed simultaneously and / or in an alternating order. For example, optional actions 1010 and 1011, and / or actions 1012 and 1013 may be performed before actions 1008 and 1009.

[0046] Action 1001

[0047] In action 1001 , the background blurring system 1 , for example with the support of the first recognition unit 101 , identifies in a first image of a scene at a first point in time at least a first pixel comprised in the background of the first image.

[0048] Action 1002

[0049] In action 1002 , the background blurring system 1 , for example with the support of the first blurring unit 102 , determines a blurred pixel value of at least a first pixel.

[0050] Action 1003

[0051] In action 1003 , the background blurring system 1 , for example with the support of the first providing unit 103 , provides blurred pixel values ​​in the video compression stream after a first point in time.

[0052] Action 1004

[0053] In action 1004 , the background blurring system 1 , for example with the support of the second recognition unit 104 , recognizes at a second later point in time in the second image of the scene that at least the first pixel has changed to be included in the foreground of the second image.

[0054] Action 1005

[0055] In action 1005 , the background blurring system 1 , for example with the support of the second providing unit 105 , continuously and / or intermittently provides the identified non-blurred pixel value of at least the first pixel in the video compression stream after the second point in time.

[0056] Action 1006

[0057] In action 1006, the background blur system 1, for example with the support of the third identification unit 106, identifies at a third subsequent time point that at least a first pixel in a third image of the scene has been changed to be included in the background of the third image, wherein the value of at least the first pixel remains unchanged from the intermediate time point until the third time point.

[0058] Action 1007

[0059] In action 1007 , when the third time point is within a predeterminable background merging time period from the middle time point, the background blur system 1 , for example with the support of the third providing unit 107 , provides blurred pixel values ​​in the video compression stream after the third time point.

[0060] Action 1008

[0061] In optional action 1008 , when the third time point is not within the background merging time period, the background blurring system 1 determines an updated blurred pixel value of at least the first pixel, for example with the support of the optional second blurring unit 108 .

[0062] Optionally, the determining of act 1008 may include additionally determining updated blurred pixel values ​​for at least additional pixels proximate to the first pixel.

[0063] Action 1009

[0064] In optional action 1009 that may follow optional action 1008, when the third time point is not within the background merging time period, the background blur system 1, for example with the support of an optional fourth providing unit 109, provides updated blurred pixel values ​​in the video compression stream after the third time point.

[0065] Optionally, the providing of action 1009 may include additionally providing updated blurred pixel values ​​for additional pixels.

[0066] Action 1010

[0067] In optional action 1010, when it is determined that the blurred pixel value of a predetermined supplementary time point has not been updated within a predetermined supplementary time period, the background blur system 1, for example with the support of an optional supplementary blur unit 110, determines an updated blurred pixel value of at least the first pixel from the current image of the scene.

[0068] Action 1011

[0069] In optional action 1011 which may follow optional action 1010, when it is determined that the blurred pixel value of a predetermined supplementation time point has not been updated within a predetermined supplementation time period, the background blur system 1 provides an updated blurred pixel value in the video compression stream after the supplementation time point, for example with the support of an optional supplementation providing unit 111.

[0070] Action 1012

[0071] In optional action 1012, when a predetermined upcoming frame of the video compression stream at a predetermined I-frame-related time point is an I-frame, the background blur system 1, for example with the support of the optional I-frame blur unit 112, determines updated blurred pixel values ​​of at least a first pixel and a predetermined portion of other pixels out of the pixels identified as being included in the background of the current image of the scene.

[0072] Action 1013

[0073] In optional action 1013 which may follow optional action 1012, when a predeterminable upcoming frame of the video compression stream at a predeterminable I-frame related time point is an I-frame, the background blur system 1 provides updated blurred pixel values ​​in the video compression stream after the I-frame related time point, for example with the support of the optional I-frame providing unit 113.

[0074] Those skilled in the art will appreciate that the present disclosure is by no means limited to the preferred embodiments described above. On the contrary, within the scope of the appended claims, many modifications and variations are possible. It should also be noted that the drawings are not necessarily drawn to scale, and that the dimensions of certain features may have been exaggerated for clarity. Instead, emphasis is placed on the principles of the embodiments described herein. In addition, in the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.

Claims

1. A method for providing a video compression stream from a camera (2) adapted to capture a scene, performed by a background blurring system (1), the method comprising: At a first point in time, identifying (1001) in a first image of the scene at least a first pixel included in a background of the first image; determining (1002) a blurred pixel value of the at least first pixel; After the first time point, providing (1003) the blurred pixel value in the video compression stream; at a later second point in time, identifying (1004) in a second image of the scene that the at least first pixel has changed to be included in a foreground of the second image; After the second time point, continuously and / or intermittently providing (1005) the identified non-ambiguous pixel value of the at least first pixel in the video compression stream; identifying (1006) in a third image of the scene that the at least first pixel has changed to be included in the background of the third image at a third point in time after the at least first pixel changes from a foreground pixel to a background pixel again; as well as When the third time point is within a predeterminable background merging time period from an intermediate time point after the second time point, and the value of the at least first pixel remains unchanged since the intermediate time point: After the third point in time, the blurred pixel value is provided (1007) in the video compression stream.

2. The method according to claim 1, further comprising: When the third time point is not within the background merging time period: determining (1008) an updated blurred pixel value for the at least first pixel; as well as After the third time point, the updated blurred pixel value is provided (1009) in the video compression stream.

3. The method of claim 2, wherein the determining (1008) the updated fuzzy pixel value comprises additionally determining the updated fuzzy pixel values ​​of additional pixels proximate to the at least first pixel, and wherein the providing (1009) the updated fuzzy pixel value comprises additionally providing the updated fuzzy pixel values ​​of the additional pixels.

4. The method according to any one of claims 1 to 3, further comprising: When the blurred pixel value at the predeterminable supplementation time point is determined to be not updated within the predeterminable supplementation time period: determining (1010) an updated blurred pixel value for the at least first pixel from a current image of the scene; and After the supplementation time point, the updated blurred pixel value is provided (1011) in the video compression stream.

5. The method according to any one of claims 1 to 3, further comprising: When a predeterminable upcoming frame of the video compression stream at a predeterminable I-frame related time point is an I-frame: determining (1012) updated blurred pixel values ​​for the at least first pixel and a predeterminable portion of other pixels out of pixels identified as being included in a background of a current image of the scene; as well as The updated blurred pixel value is provided (1013) in the video compression stream after the I-frame related time point.

6. The method according to any one of claims 1 to 3, wherein the kernel size of the image filter kernel used to determine the blurred pixel value of the at least first pixel matches the macroblock size of the I frame of the video compression stream.

7. A background blurring system (1) for providing a video compression stream from a camera (2) adapted to capture a scene, the background blurring system comprising: A first recognition unit (101) for recognizing (1001) at a first time point in a first image of the scene at least a first pixel included in a background of the first image; a first blurring unit (102) for determining (1002) a blurred pixel value of the at least first pixel; A first providing unit (103) for providing (1003) the blurred pixel value in the video compression stream after the first time point; a second recognition unit (104) for recognizing (1004) in a second image of the scene at a later second point in time that the at least first pixel has changed to be included in the foreground of the second image; a second providing unit (105) for continuously and / or intermittently providing (1005) the identified non-ambiguous pixel value of the at least first pixel in the video compression stream after the second time point; a third identification unit (106) for identifying (1006) in a third image of the scene that the at least first pixel has changed to be included in the background of the third image at a third time point after when the at least first pixel changes from a foreground pixel to a background pixel again; as well as A third providing unit (107) is used to provide (1007) the blurred pixel value in the video compression stream after the third time point when the value of at least the first pixel remains unchanged starting from the intermediate time point within a predeterminable background merging time period from the intermediate time point after the second time point.

8. The background blurring system (1) according to claim 7, further comprising: a second blurring unit (108) for determining (1008) an updated blurred pixel value of the at least first pixel when the third time point is not within the background merging time period; as well as A fourth providing unit (109) is used to provide (1009) the updated blurred pixel value in the video compression stream after the third time point.

9. A background blur system (1) according to claim 8, wherein the second blurring unit (108) is suitable for additionally determining updated blurred pixel values ​​of additional pixels near the at least first pixel, and wherein the fourth providing unit (109) is suitable for additionally providing the updated blurred pixel values ​​of the additional pixels.

10. The background blurring system (1) according to any one of claims 7 to 9, further comprising: a supplementary blurring unit (110) for determining (1010) an updated blurred pixel value of the at least first pixel from a current image of the scene when the blurred pixel value at the predeterminable supplementary time point is determined to have not been updated within a predeterminable supplementary time period; as well as A supplementation providing unit (111) is used to provide (1011) the updated blurred pixel value in the video compression stream after the supplementation time point.

11. The background blurring system (1) according to any one of claims 7 to 9, further comprising: an I-frame blurring unit (112) for determining (1012) updated blurred pixel values ​​for said at least first pixel and a predetermined portion of other pixels out of pixels identified as being included in a background of a current image of said scene when a predetermined upcoming frame of said video compression stream at a predetermined I-frame related time point is an I-frame; as well as An I frame providing unit (113) is used to provide (1013) the updated blurred pixel value in the video compression stream after the I frame related time point.

12. A background blur system (1) according to any one of claims 7 to 9, wherein the kernel size of the image filter kernel used to determine the blurred pixel value of the at least first pixel matches the macroblock size of the I frame of the video compression stream.

13. A camera (2) comprising a background blurring system (1) according to any one of claims 7 to 12.

14. A computer program product stored on a computer readable medium comprising a computer program arranged to cause a computer or a processor to perform the steps of the method according to any one of claims 1 to 6.

15. A non-volatile computer-readable storage medium having stored thereon the computer program product according to claim 14.

Images