Circuitry and method

Abstract

A circuitry for diagnosing external interference influencing a monitoring image captured with a monitoring camera is configured to obtain reference intensities of reference pixels of a reference marker region in a reference image captured with the monitoring camera, wherein the reference marker region is smaller than the reference image, obtain the monitoring image, determine current intensities of sample pixels of a marker region in the monitoring image, wherein the marker region corresponds to the reference marker region, and wherein the sample pixels of the marker region corresponds to the reference pixels of the reference marker region, determine a correlation between the determined current intensities and the reference intensities, and diagnose, based on the determined correlation and a predefined threshold, the monitoring image for degradation due to an external interference for evaluating the reliability of the monitoring camera as a monitoring tool.

Description

[0001] Our ref.: 240259EPWOP 1Sony Semiconductor Solutions Corporation CIRCUITRY AND METHODTECHNICAL FIELDThe present disclosure generally pertains to a circuitry and a method.TECHNICAL BACKGROUNDGenerally, it is known that monitoring is important in many domains, such as domains that relyon the captured images for critical decision making. For example, where it is desirable tomonitor hardware or optical performance, e.g., in the security or automotive domain or whenimage capture is used as legal evidence. Also, for example, in monitoring where heavymachinery is used, and workers could come into close proximity of the machinery, a camera-based surveillance system can be used to ensure the safety of workers. The monitoring systemcan be designed to continuously or in repeated intervals monitor an area of interest, for examplethe area surrounding machinery, and to analyze visual data, for example, to detect the presenceof personnel within predefined danger zones. In some cases, also appropriate alerts, for example,in case of a detected worker coming dangerously close to a machine, can be provided. Thereby,undesirable incidents, accidents and injuries, for example, of workers that inadvertently come tooclose to hazardous machinery, can be prevented.Similarly, for example, when monitoring an environment with children or for assisted living,e.g., monitoring of people with care requirements, critical decision making, for example, regarding safety or when deciding if assistance is required, often relies on monitoring. Although there exist techniques for monitoring, it is generally desirable to improve on existing techniques. SUMMARYAccording to a first aspect the disclosure provides circuitry for diagnosing external interferenceinfluencing a monitoring image captured with a monitoring camera, configured to:obtain reference intensities of a plurality of reference pixels of a reference marker region in a reference image captured with the monitoring camera, wherein the reference marker region is smaller than the reference image, obtain the monitoring image, determine current intensities of a plurality of sample pixels of a marker region in the monitoring image, wherein the marker region corresponds to the reference marker region, and wherein the plurality of sample pixels of the marker region corresponds to the plurality of reference pixels ofOur ref.: 240259EPWOP 2Sony Semiconductor Solutions Corporation the reference marker region, determine a correlation between the determined current intensities and the reference intensities, and diagnose, based on the determined correlation and a predefined threshold, the monitoring image for degradation due to an external interference for evaluating the reliability of the monitoringcamera as a monitoring toolAccording to a second aspect the disclosure provides a method for diagnosing externalinterference influencing a monitoring image captured with a monitoring camera, comprising:obtaining reference intensities of a plurality of reference pixels of a reference marker region in a reference image captured with the monitoring camera, wherein the reference marker region is smaller than the reference image, obtaining the monitoring image, determining current intensities of a plurality of sample pixels of a marker region in the monitoring image, wherein the marker region corresponds to the reference marker region, andwherein the plurality of sample pixels of the marker region corresponds to the plurality ofreference pixels of the reference marker region, determining a correlation between the determined current intensities and the reference intensities, and diagnosing, based on the determined correlation and a predefined threshold, the monitoringimage for degradation due to an external interference for evaluating the reliability of themonitoring camera as a monitoring tool.Further aspects are set forth in the dependent claims, the drawings and the following description.BRIEF DESCRIPTION OF THE DRAWINGS Embodiments are explained by way of example with respect to the accompanying drawings, in which:Fig. 1 shows a table of reliability values and the diagnostic indicators correlation, mean whiteintensity, mean black intensity, contrast and blur according to an embodiment; Fig.2 shows an example of a reference marker region and marker region defined by virtual markers and selection of pixels based on the virtual marker layout according to an embodiment;Fig. 3 shows an example of a real marker defining the reference marker region and the markerregion according to an embodiment;Our ref.: 240259EPWOP 3Sony Semiconductor Solutions CorporationFig. 4 shows an embodiment for diagnosing external interference influencing a monitoring imagecaptured with a monitoring camera;Fig. 5 shows an embodiment for diagnosing external interference influencing a monitoring imagecaptured with a monitoring camera using multiple diagnostics;Fig. 6 shows an embodiment of a method for diagnosing external interference influencing amonitoring image captured with a monitoring camera;Fig. 7 shows an example of an initialization process for determining reference parametersaccording to an embodiment;Fig. 8 shows an example of a contrast check of a reference marker region according to anembodiment; and Fig.9 shows a block diagram depicting an embodiment of an electronic device for diagnosingexternal interference influencing a monitoring image captured with a monitoring camera.DETAILED DESCRIPTION OF EMBODIMENTSBefore a detailed description of the embodiments under reference of Figs. 1 to 9 is given, generalexplanations are made.As explained in the outset, monitoring based on camera surveillance is crucial for preventingnegative incidents, accidents and injuries, for example, of workers that could come into contactwith heavy machinery (industrial safety monitoring), or for example incidents regarding securityor care concerns. It has been recognized that it is also critical to constantly monitor the quality of the imagesproduced by the monitoring camera. That is, the images produced by the monitoring cameraneed to be of high enough quality to allow the incident detection algorithm, e.g., the system that analyzes the images for undesirable or dangerous situations (e.g., a person entering a predefined danger zone, a person needing assistance, a security incident, etc.), to operate reliably. In other words, it has been recognized that diagnostics of the monitoring images are needed.In particular, it has been recognized that external interference influencing the monitoring imagescaptured with the monitoring camera, such as dirt on the lens or dirt and / or smoke in the air candegrade the monitoring image captured by the monitoring camera.Thus, some embodiments pertain to a circuitry for diagnosing external interference influencing amonitoring image captured with a monitoring camera, configured to:Our ref.: 240259EPWOP 4Sony Semiconductor Solutions Corporation obtain reference intensities of a plurality of reference pixels of a reference marker region in a reference image captured with the monitoring camera, wherein the reference marker region is smaller than the reference image, obtain the monitoring image, determine current intensities of a plurality of sample pixels of a marker region in the monitoring image, wherein the marker region corresponds to the reference marker region, and wherein the plurality of sample pixels of the marker region corresponds to the plurality of reference pixels of the reference marker region, determine a correlation between the determined current intensities and the reference intensities, and diagnose, based on the determined correlation and a predefined threshold, themonitoring image for degradation due to an external interference for evaluating the reliability ofthe monitoring camera as a monitoring tool.Circuitry The circuitry may include one or more processors, logical circuits, memory (read only memory, random memory, etc., storage, e.g., hard disk, compact disc, flash drive, etc.), an interface for (wireless, e.g., Bluetooth, infrared) communication via a network (local area network, wireless network, internet). Moreover, it may include input means (mouse, keyboard, microphone, camera etc.), output means (loudspeakers, display (e.g., liquid crystal, (organic) light emitting diode, etc.)), and one or more sensors for sensing still image or video image data (image sensor,camera sensor, video sensor, etc.). The sensor and / or input means may include the monitoringcamera. The circuitry may also include sensor(s) for sensing audio data (microphone).The memory may store the monitoring image and / or reference image and / or the respective pixelinformation of the reference pixels and / or the sample pixels, such as reference intensities or current intensities. Also, the circuitry may receive, for example, via the interface, the reference image and / or the reference marker region and / or reference pixel information from an outside source, for example, from a server, or for example, from the monitoring camera.Our ref.: 240259EPWOP 5Sony Semiconductor Solutions CorporationSimilarly, the circuitry may receive, for example, via the interface, the monitoring image and / orthe marker region and / or sample pixel information from an outside source, for example, from the monitoring camera. External interference External interference, e.g., from outside the monitoring camera, such as, from the environment,(e.g., the weather, users, other devices or machines etc.), influencing the monitoring imagecaptured with the monitoring camera, e.g., leading to deterioration in camera performance, mayinclude an interference that occludes or partially occludes the camera lens.For example, particles, which may be airborne, and / or residues on the lens or in the air (e.g., infront of the lens), such as dirt (e.g., calcium deposits where water has dried, dust, grease, etc.),smoke, water (e.g., vapour, condensation, frost, raindrops or the like), etc., can degrade themonitoring image captured by the monitoring camera. But external interference may also include mechanical shock or overheating, both of which may cause the lens of the monitoring camera to lose its focus, loss illumination (e.g., because of lightsource failure), excess illumination (e.g., because of direct daylight) or the like.The deterioration due to external interference may occur gradually over time.External interference may occur in a manner that renders the monitoring camera unreliable forthe purpose of monitoring as described above.In this way, image diagnostics may include a verification on whether the monitoring camera is interfered with externally, for example, occluded, partially or completely. Monitoring camera The monitoring camera may be any type of camera, for example, an RGB camera, a depth sensing camera (e.g., dToF (Direct Time-of-Flight) and iToF (Indirect Time-of-Flight) camera),an event-based camera or the like. The monitoring camera may be static or movable. Themonitoring camera may include one or multiple lenses and / or one or multiple sensors, forexample a stereoscopic camera (left and right), which may be in one housing. Also, themonitoring camera may include multiple cameras, or multiple independent cameras with overlapping field of views. In the case of a stereoscopic camera, the left and right images of the respective left and right stereoscopic lenses may be compared for diagnostics, for example for determining the location of the reference pixels of the reference image in relation to the monitoring image for selectingOur ref.: 240259EPWOP 6Sony Semiconductor Solutions Corporation the sample pixels from the monitoring image, so that they correspond to the reference pixels. The same may apply in case of multiple independent cameras or moving cameras. Reference Image The reference image may be a color or a grayscale image of the scene. The scene may includethe monitoring area of interest, for example, the machinery and a predefined danger zone aroundthe machinery, that is surveilled.The reference image may be captured by the same monitoring camera that also captures the monitoring image, for example if the monitoring camera is static. The capture of the reference image may occur prior to the capture of the monitoring image in an initialization process. In this initialization process (for example, occurring at installation when the scene is cleared), information of the reference marker region may be extracted using manually provided centercoordinates and sizes (e.g., radius and / or diameter), for example if the reference marker region isbased on a real marker as explained below in more detail.For stereoscopic cameras, which may, for example, compare left and right images as referenceimage and monitoring image, or for independent cameras with overlapping field of views,wherein the reference image and the monitoring image may be captured by the independentcameras, or for moving cameras, a registration between the reference image and the monitoring image may be performed, in such a way that the position of the reference pixels with respect to the monitoring image is determined. In this way, sample pixels can be selected from the monitoring image that correspond to the reference pixels. With respect to the moving camera, the registration may be performed by a SLAM (Simultaneous Localization and Mapping) algorithm, which may compare a previous referenceimage with a current monitoring image. Also, if the monitoring camera, includes a depth sensingcamera, the registration between the reference image and the monitoring image may be based on the depth measurement. For example, this may also apply if the depth sensing camera is a moving camera. Reference Marker region The reference marker region may be defined during an initialization process. The reference marker region is smaller than the reference image.Thus, it may be smaller than the field of view of the monitoring camera displayed in thereference image in at least one direction.Our ref.: 240259EPWOP 7Sony Semiconductor Solutions Corporation The reference marker region may include multiple reference marker regions. Thus, diagnosticsmay be based on multiple reference marker regions and therefore, may also be based, on multiplemarker regions. In this way, even small external interferences, which would otherwise notimpact a single marker region in the monitoring image, may be detected.In some embodiments, the reference marker region may have a reliable contrast, whereinreliability of the contrast may be based on a Pearson correlation coefficient of a Pearsoncorrelation between the current intensities of the plurality of sample pixels and the referenceintensities of the corresponding plurality of reference pixels being above a predefined contrastthreshold. The Pearson correlation coefficient may be determined as described below regarding the formula “^^^^^^^” of the Pearson correlation coefficient and its described parameters.The contrast threshold may be at least 0.9, for example 0.9 or between 0.9 and 1.If the determined Pearson correlation coefficient is below the threshold, e.g., below 0.9, it maybe determined that the reference marker region does not have sufficient contrast, i.e., hasunreliable contrast. In that case, another reference marker region may be selected. For example,another real or virtual marker may be selected as defining the reference marker region, asexplained in more detail below.Also, the contrast may be the greyscale contrast or the color contrast, depending on the type ofintensity that is determined as described below.The color contrast may, for example, take into account the difference (contrast) in colors and / orthe Y (luminance) of the YUV color model, wherein the greyscale contrast may take into account the luminance of the greyscale image, as described below. Real Marker The reference marker region may be defined by a real marker placed in the imaged scene. Thatis, during the set-up of the monitoring camera, real markers may be placed in the scene, whichmay then be indicated during initialization of the reference marker region. At least one markermay be placed to have no horizontal, vertical or central symmetry in the image, because in this way sensor failure, for example, sensor hardware failure (e.g., failures that cause the image to flip) or sensor failure concerning the image width and offset may be detected. The relevant areas of the real marker may be stored in (persistent) memory after initialization orreused for every monitoring image to determine the diagnostics as described below. Also, a resetOur ref.: 240259EPWOP 8Sony Semiconductor Solutions Corporation concerning the marker and relevant marker areas may be provided for the monitoring camera for reinstallation and / or re-initialization. The multiple reference marker regions may be defined by multiple real markers placed in the scene.The real marker may include predefined black and white parts, which ensures sufficient contrastfor subsequent diagnostics. However, also other color contrasts, for example red and green,contrasts may be possible. Thus, the real marker may include predefined transitions betweenblack to white parts (or between any other colors of the real marker). For example, some parts of the real marker may be predefined as a transition location, where pixels are close according to a predefined threshold to the transition from one color to the other color (e.g., from black to white). Some parts of the real marker may be predefined as a high frequency area, where variation, for example black and white variation, occurs in fewer pixels, compared to a low frequency area, which may be predefined as an area with less or no variation, for example a wider area (with more pixels) of the same color (e.g., black or white). Some parts of the real marker may be predefined as safe location, where pixels are far, according to a predefined threshold, from any transition location.Also, the real marker may be designed so that degradation of the monitoring image will result inunderestimates of the current parameters used in diagnostics. Virtual Marker The reference marker region may be defined by a virtual marker in the scene, for example by a static part present in the scene. That is, a region of the reference image that represents a static part of the imaged scene captured in the reference image may define the reference marker region.For example, any part of the image, which presents a contrast, may be used as a virtual marker,wherein a higher contrast provides improved contrast diagnostics (see below), e.g., contrast change detection, and including sharper transitions provides better sharpness loss detection (e.g., see diagnostics below). Also, edges in an image, e.g., boundary between two areas of differentcolour or brightness, such as, a (static) object's shadow on a white wall, may be included as thevirtual marker. The same monitoring camera may capture an image as a reference image prior to capturing the monitoring image and based on the prior reference image the virtual marker defining the reference marker region may be selected.Our ref.: 240259EPWOP 9Sony Semiconductor Solutions Corporation For this reason, the monitoring camera may not move between capture of the reference image and capture of the monitoring image(s). To select a virtual marker defining the reference marker image, the reference image may be split into a uniform grid of candidate virtual markers.Each virtual marker may include at least one square Region Of Interest (ROI) where a number ofreference pixels (points) may be sampled from the reference image. The reference pixels to besampled from the reference image may be arranged either as a vertical crosshair pattern centeredon the virtual marker’s center, and / or after adding diagonal lines to this pattern, may be sampledfrom the diagonal lines. For example, the crosshair pattern may span each virtual marker’s ROI, so that for example apitch of 32 will result in 65 points per virtual marker, i.e., 65 reference pixels may be sampled,and if two diagonals are used additionally to the crosshair pattern a pitch of 32 will result in 129points per marker with diagonals, i.e., 129 reference pixel may be sampled.Generating the grid and sampling the reference image may be accomplished in an initializationprocess. The multiple reference marker regions may be defined by multiple virtual markers. The multiple reference marker regions may be defined by a mixture of real markers and virtual makers.Reference and Sample Pixels, obtaining reference intensities and determining current intensitiesThe plurality of reference pixels may include a minimum number of pixels required to compute valid statistics (such as a correlation). That is, the minimum sample size of reference pixels maybe 10. However, the sample size of reference pixels may be 20 to 30.The same may apply to the plurality of sample pixels. In this way, joint statistics (e.g., correlation) may be performed.The plurality of reference pixels may be selected from the reference marker region in thereference image during an initialization (process).Furthermore, during the initialization, the reference intensities of the reference pixels may be determined and may then be stored in memory. Obtaining the reference intensities of the plurality of reference pixels may, therefore, includeobtaining the intensities from memory or from an external source (e.g., a server).Our ref.: 240259EPWOP 10Sony Semiconductor Solutions Corporation Alternatively, obtaining the reference intensities from the plurality of reference pixels may include performing the initialization process. Also, obtaining reference intensities of a plurality of reference pixels may refer to obtaining an intensity for each reference pixel of the plurality of reference pixels. For example, for each reference pixel selected during the initialization process. Similarly, determining current intensities of a plurality of sample pixels may refer to obtaining an intensity for each sample pixel of the plurality of sample pixels.In some embodiments, the intensity may be the luminance. For example, reference intensitiesmay be based on the luminance values of the plurality of reference pixels and the current intensities may be based on the luminance values of the plurality of sample pixels. The luminance of a pixel (reference or sample) may be obtained based on the greyscale image.For example, the pixels’ luminance values of the greyscale image may range from 0 to 255representing the different shades of gray.Alternatively, the intensity or luminance may be based on the color image. The intensity of thepixel may, for example, be based on at least one or all of the Y, U, V values of the YUV colormodel. For example, the intensity may be based on the Y (luminance) value. Also, determining the current intensities from the plurality of sample pixels may be performed ina corresponding manner as determining the reference intensities from the plurality of referencepixels. Monitoring imageThe monitoring image may include the static elements of the scene of the reference image. Thatis, the monitoring image scene may include the machinery and the predefined danger zone.The monitoring image may include a color or a grayscale image of the scene or both.Also, diagnostics may be performed on multiple monitoring images, for example on each of a sequence of monitoring images.For example, the monitoring image may be a frame from a surveillance video of the monitoringcamera. That is, diagnostics may be performed for each frame (i.e., for each monitoring image) of the surveillance video.Our ref.: 240259EPWOP 11Sony Semiconductor Solutions Corporation Thus, determining the intensities, determining the correlation and performing the diagnosis maybe performed for each monitoring image of the multiple monitoring images, e.g., of thesurveillance video.Obtaining the monitoring image may refer to obtaining the monitoring image from memory (oran external source) and / or may refer to capturing the monitoring image with the monitoringcamera and / or obtaining the monitoring image from the monitoring camera.For example, the monitoring camera may capture the monitoring image (and also the reference image) in advance and the monitoring image is saved (or uploaded) in a storage (e.g., memory,server, external device etc.). In this way the monitoring image may be obtained from the storage,such that the quality control of the monitoring image (including e.g., obtaining the monitoring image and / reference parameters, determining current parameters, diagnosing based on the parameters) may be performed at a later point in time, for example, when the monitoring camera is shut off. Marker region As the marker region corresponds to the reference marker region, the above explanationconcerning reference marker regions, real and virtual markers may also apply to the markerregion. For example, the marker region may include multiple marker regions in the monitoring image.That is, the marker region of the monitoring image may represent the same part of the scene,e.g., the static part of the scene, that is represented by the corresponding reference marker regionin the reference image. For example, if the reference image and the monitoring image are captured by the same static camera, the size, location and position of the reference marker region with respect to the reference image may match the size, location and position of the marker region with respect to the monitoring image. Sample Pixels As the plurality of sample pixels correspond to the plurality of reference pixels, there is a matching relationship between the reference pixels of the reference image and the sample pixelsof the monitoring image. For example, they may correspond to the same location in the real-world scene.Our ref.: 240259EPWOP 12Sony Semiconductor Solutions Corporation The sample pixels may be selected from the monitoring image such that they match the positionof the reference pixels in relation to the reference image scene. This can be accomplished basedon the marker region wich corresponds to the reference marker region. Correlation Determining the correlation (e.g., the Pearson correlation) may be based on computing thecorrelation coefficient, for example the Pearson correlation coefficient. The correlationcoefficient may indicate whether the marker region is visible or not.The Pearson correlation coefficient ^^^^^^^ may be computed according to the followingformula: wherein n is the number of samples, (i.e., in the formula above the number of reference pixels, as well asthe number of sample pixels), is the reference intensity of each reference pixel of the plurality of referencewherein ^^,^^^is the current intensity of each sample pixel of the plurality of sample pixels, wherein ^ ^^^^= ^∑^ ^^^,^^^is the mean reference intensity of the plurality of reference pixels, i.e., the mean of the reference intensities, wherein ^ ^^^^= ^∑^ ^^^,^^^ is the mean current intensity of the plurality of sample pixels, i.e., the meanof the current intensities, wherein^ the standard deviation of the reference intensities of plurality of reference pixels, wherein ^ = is the standard deviation of the current intensities of the of sample pixels.In some embodiments, determining the correlation may include determining a noise-adjustedPearson correlation coefficient ^^^^ according to the following formula:Our ref.: 240259EPWOP 13Sony Semiconductor Solutions Corporation wherein ^^^^ and ^^^^ are predefined correction coefficients and wherein the other parameters of the^^^^ formula are described above regarding the Pearson correlation coefficient ^^^^^^^.For example, ^^^^ may be a predefined correction coefficient depending on the noise in theintensities of the plurality of reference pixels, and ^^^^ may be a predefined correctioncoefficient depending on the noise in the intensities of the plurality of sample pixels.In this way, the noise-corrected Pearson formula ^^^^ can correct the ^^^^^^^ with a term thatmay depend on the noise in the pixel intensity.For example, when using a low contrast real or virtual marker for defining the reference markerregion and marker region, in particular, when using a low contrast virtual marker, for each pixelthe noise of the pixels may degrade the normal Pearson correlation coefficient. However, usingthe noise-adjusted Pearson correlation coefficient ^^^^ allows recognition and usage of lowcontrast markers.As explained above, to determine whether the contrast computed for such low contrast makers isreliable, the Pearson correlation coefficient “^^^^^^^” may be computed and compared to apredefined contrast threshold (e.g., 0.9).However, if the Pearson correlation coefficient “^^^^^^^” falls below the contrast threshold, themarker may be too unreliable to be used as a reference marker region and another marker, forexample a higher contrast virtual marker, may be selected as the reference marker region and thecorresponding marker region. Diagnostics In the presence of certain degradations of the monitoring image, several diagnostic indicators and / or several current parameters of the monitoring image may become inaccurate, for example,as compared to the reference parameters of the reference image (e.g., the diagnostic indicatorsmay compare current parameters of the monitoring image with corresponding reference parameters of the reference image), which is explained in the following in more detail. Diagnostic indicators may be the intensities as well as the associated correlation of intensities,contrast (e.g., the contrast between the white and black parts of the marker region), blur, themean white intensity, e.g., the mean intensity of the white parts of the marker region, and / or the mean black intensity, e.g., the mean intensity of the black parts of the marker, which are explained in more detail below.Our ref.: 240259EPWOP 14Sony Semiconductor Solutions Corporation Furthermore, diagnostic indicators may be the saturation in the white part of the marker region, ifany, or the sharpness of transitions (e.g., white / black transitions).Occlusion For example, during occlusion, the current intensities may not reflect the ambient light anymore, which may be diagnosed based on the correlation.Thus, the correlation coefficient may be a diagnostic indicator for occlusion.Diagnosing the monitoring image for degradation may, therefore, include comparing thecomputed correlation coefficient with the predefined threshold.If the correlation coefficient falls below the predefined threshold the marker region may bedetermined to be occluded. Alternatively, based on this result in combination with otherdiagnostics, the marker region may be determined to be occluded (e.g., contrast failure see below). If the marker has been determined to be occluded, the image may be diagnosed to be degraded. Alternatively, only in combination with other diagnostics, the image may be diagnosed to be degraded. If the image is diagnosed to be degraded, the monitoring camera may be evaluated to be unreliable for monitoring. In case the monitoring camera is evaluated to be unreliable for monitoring, an appropriate alert or notification may be sent (for example to an operator, observer, or to a worker of the machinery).Additionally, also contrast may, during occlusion, not reflect the ambient light anymore, whichmay be diagnosed as described below in section additional parameters and diagnostics. Therefore, contrast may also be a diagnostic indicator for occlusion. Thus, for example, the correlation in combination with the contrast may be a diagnostic indicator for occlusion, as during occlusion both, the current intensities and the contrast may not reflect ambient light anymore. Also, in case multiple marker regions are used, the image may only be diagnosed to be degradedif it has been determined that the occluded marker regions are above a predefined threshold. Forexample, the monitoring image may be diagnosed to be degraded if above 50% of the marker regions are determined to be occluded.Our ref.: 240259EPWOP 15Sony Semiconductor Solutions Corporation Also, certain marker regions may be classified as belonging to a monitoring region of interest (e.g., a danger zone around a machinery). Thus, the image may be diagnosed to be degraded based on the classification of marker regions, for example, if the occluded marker region is classified as including a monitoring region of interest, the image may be diagnosed to be degraded.In some embodiments, the monitoring image may be diagnosed as degraded and the monitoringcamera as unreliable for monitoring if the noise-adjusted Pearson correlation coefficient ^^^^ islower than a predefined threshold ^^^^^^^.In the above ways, image diagnostics may include a verification on whether the camera orcamera lens is occluded, partially or completely (for example, if dirt or smoke interfereexternally with the monitoring) and it may also include a verification on whether the occlusionlies in a monitoring region of interest or outside of a monitoring region of interest.Additional degradations and diagnosticsUnder / Overexposure, contrast & blurFor example, it has been recognized that blur can decrease the current intensities close to the color transitions and that substantial blur can make the transition hard to distinguish from noise.Furthermore, additional diagnostics may also verify that the camera scene is not under oroverexposed, such as the intensity ratio between the current intensities and the reference intensities. For example, if the intensity ratio is lower than a predefined threshold the monitoring image may be diagnosed to be underexposed and if the intensity ratio is higher than a predefined threshold the monitoring image may be diagnosed to be overexposed. Moreover, diagnostics may also verify whether the lens is out of focus or not.Thus, in some embodiments, the circuitry may be further configured to determinea mean white intensity of the plurality of sample pixels, and / ora mean black intensity of the plurality of sample pixels, and / ora contrast of the plurality of sample pixels, and / ora blur of the plurality of sample pixels,wherein diagnosing the monitoring image may be further based onthe mean white intensity of the plurality of sample pixels, and / or the mean black intensity of the plurality of sample pixels, and / orOur ref.: 240259EPWOP 16Sony Semiconductor Solutions Corporation the mean black intensity of the plurality of sample pixels, and / or the blur of the plurality of sample pixels.The mean white intensity may refer to the average intensity, for example the average YUVvalue, of white parts of the marker region.The mean black intensity may refer to the average intensity, for example the average YUV value,of black parts of the marker region. The location of the white and black parts within the markerregion may be determined, for example, based on information of the predefined reference markerregion. For example, the location of the white and black parts, i.e., which pixels of the plurality of sample pixels fall in a white or black part, may be obtained based on the reference marker region, which may be predefined, for example by selecting a real marker with black and whiteparts for defining the reference marker region.The contrast may refer to the Michelson contrast, defined as: wherein ^^^^^^ is the intensity of all sample pixels of the white part of the marker region, andwherein ^^^^^^ is the intensity of all sample pixels of the black part of the marker region.The blur may refer to the number of sample pixels needed to switch from white to black in thetransition between predefined black and white parts of the marker region.Afterwards, the mean white intensity of the plurality of sample pixels may be compared to apredefined white intensity threshold.Similarly, the mean black intensity of the plurality of sample pixels may be compared to apredefined black intensity threshold.The contrast of the plurality of sample pixels may be compared to a predefined contrastthreshold.The blur of the plurality of sample pixels may be compared to a predefined blur threshold.For example, if the mean white intensity, the mean black intensity and / or the contrast fall below their respective predefined threshold and / or the blur is above its respective predefined blurthreshold, the monitoring image may be determined to be degraded. This may occur for exampledue to smoke or light scattering.Furthermore, any of the thresholds may be based on obtained reference parameters of thereference marker.Our ref.: 240259EPWOP 17Sony Semiconductor Solutions CorporationThat is, the mean white intensity, the mean black intensity, the contrast and the blur determinedbased on the sample pixels may be compared to corresponding reference mean white intensity,reference mean black intensity, reference contrast and reference blur of the reference pixels. Allthe reference parameters may be computed as described for determining the current parameters of the monitoring image. For example, the contrast may refer to the Michelson contrast, defined as: ^^^ℎ^^^^^ ^^^^^^^^ =^^^^^^^^^^^^^^^^^^^^^^^^^^, with corresponding parameters as described above. In some embodiments, the circuitry may be further configured to obtaina mean white intensity of the plurality of reference pixels, and / ora mean black intensity of the plurality of reference pixels, and / ora contrast of the plurality of reference pixels, and / ora blur of the plurality of reference pixels,wherein the circuitry may be further configured to determinea mean white intensity comparison between the mean white intensity of the plurality of samplepixels and the mean white intensity of the plurality of reference pixels, and / ora mean black intensity comparison between the mean black intensity of the plurality of samplepixels and the mean white intensity of the plurality of reference pixels, and / or a contrast ratio between the contrast of the plurality of sample pixels and the contrast of the plurality of reference pixels, and / or a blur ratio between the blur of the plurality of sample pixels and the blur of the plurality of reference pixels,wherein diagnosing the monitoring image may be based on the mean white intensity ratio, and / orthe mean black intensity ratio, and / or the contrast ratio, and / or the blur ratio.Furthermore, the contrast ratio may be determined according to^^^^^^^^^^^^^^^^^^^^^^, wherein the blurratio may be determined according to wherein wherein wherein Our ref.: 240259EPWOP 18Sony Semiconductor Solutions Corporation ^ ^^^^= ^∑^ ^^^,^^^is the mean current intensity of the plurality of sample pixels, wherein ^^^^ is the mean reference intensity of the plurality of reference pixels, wherein ^ the standard deviation of the current intensities, ^ is the standard deviation of the reference intensities, ^^ ^^^^^^^=^^^^^^^,^^^, wherein^^^^,^^^ is the maximum difference in current intensities between adjacent pixels of the pluralityof sample pixels, wherein ^^ ^^^^^^^=^^^^^^^,^^^, wherein^^^^,^^^ is the maximum difference in reference intensities between adjacent pixels of theplurality of reference pixels. Also, the mean white intensity of the plurality of sample pixels may be determined according to: ^^^^^ℎ^^^^^^ = ^^^^ + ^^^^The mean black intensity of the plurality of sample pixels may be determined according to: ^^^^^^^^^^^^ = ^^^^ − ^^^^The mean white intensity of the plurality of reference pixels may be determined according to: ^^^^^ℎ^^^^^^ = ^^^^ + ^^^^The mean black intensity of the plurality of reference pixels may be determined according to: If the contrast ratio is lower than a predefined threshold ^^^^^^^^^^^^^^^^, the marker regionmay be determined to fail the contrast check. On this basis, the monitoring image may be diagnosed as degraded and the monitoring camera as unreliable for monitoring.If the blur ratio is larger than a predefined threshold ^^^^^^^^^^^^, the marker region may bedetermined to fail the blur check. On this basis, the monitoring image may be diagnosed asdegraded and the monitoring camera as unreliable for monitoring.Our ref.: 240259EPWOP 19Sony Semiconductor Solutions Corporation If the mean black intensity comparison is outside of predefined tolerances, the marker region may be determined to fail the mean black intensity check. On this basis, the monitoring imagemay be diagnosed as degraded and the monitoring camera as unreliable for monitoring.If the mean white intensity comparison is outside of predefined tolerances, the marker region may be determined to fail the mean white intensity check. On this basis, the monitoring imagemay be diagnosed as degraded and the monitoring camera as unreliable for monitoring.The same may apply to any of the other diagnostic indicators and any combination of the diagnostic indicators described in this specification. That is, the diagnosis of degraded or undegraded monitoring image (i.e., of reliable or unreliable monitoring camera for monitoring) may be based on any combination of results of the diagnostics, e.g., on any combination of failed or passed checks.Shape diagnostics and substantial blurMoreover, diagnosing the degradation of the monitoring image may be based on determiningwhether the shape and position of the marker region corresponds to the shape and position of the reference marker region, for example if a real marker is used (see Fig.3, in which the outer circles 3,4 of real marker 1 are thinner than the central and second circles 5,2). In this way, substantial blur may be determined if the difference in shape and / or position is above apredefined threshold.Also, any combination of the diagnostics and / or diagnostic indicators described in this specification may be used to diagnose the degradation of the monitoring image. The above-described diagnostic indicators may be computed using pixels that fall in safe areas (i.e., far enough from transitions). Reliability value Moreover, for diagnosing the image to be degraded or not degraded a reliability value may be determined. For example, the reliability value may function as a reliability flag, which may beset to unreliable if degraded conditions are observed. Also, some diagnostic indicators, forexample, detection of saturation may trigger the reliability value to be unreliable.Thus, the reliability value of the marker region of the monitoring image may depend on any ofthe diagnostics described in this specification.Also, if the reliability value indicates unreliable, e.g., in a degraded case, other diagnosticindicators of that marker region may be determined to be unreliable and may be discarded.Our ref.: 240259EPWOP 20Sony Semiconductor Solutions Corporation In other words, the reliability value may indicate the reliability of diagnostic indicators. For example, the reliability value may be determined based on the diagnostic indicator correlationand may in turn give an indication if other diagnostic indicators (for example if other currentparameters) e.g., mean white intensity, mean black intensity, contrast and / or blur, are degraded or not. Determining the reliability value may be based on sequential checks of any one or more of the above diagnostics including, for example, determining the location, shape and movement of the marker region in the monitoring image compared to the reference marker region. Implementation examples In the following, an implementation example with six checks is explained. In this example ameasurement is marked as “Unreliable” when some degraded conditions are observed. Theunreliability reason may be detailed through an enumeration. Determining the reliability valuemay be the result of sequential checks (see the Order) according to the following example:Order 1. If, the correlation is smaller than a threshold, the reliability value may indicate that thecorrelation is too low (e.g., see “Too low correlation”, Fig. 1).Order 2. If, the mean white intensity is smaller than a minimal contrast ratio, the reliability valuemay indicate that the marker region is too dark (e.g., see “Too dark”, Fig. 1).Order 3. If, the mean black intensity is bigger than a maximum black intensity, the reliabilityvalue may indicate that the marker region is too bright (e.g., see “Too bright”, Fig. 1).Order 4: If, the contrast of a high frequency area in the marker region is below the minimal high frequency contrast or the ratio of the contrasts on high or low frequency areas is below a minimum contrast ratio, the reliability value may indicate that the marker region is too blurred (e.g., see “Too blurred”, Fig.1). Order 5: If, saturation is found in the sample pixels, the reliability value may indicate that thesample pixels are saturated (e.g., see “Saturated”, Fig. 1).Order 6: For other reasons, for example, if the blur measurement is not successful, e.g., a high level of noise in a low luminosity monitoring image occurred, the reliability value may be set to“Unreliable” (e.g., see “Unreliable” column 2, Fig. 1).If nothing is wrong, e.g., none of the previous Order 1 to 6 conditions are met, the reliabilityvalue may be set to “Reliable” (e.g., see “Reliable”, Fig. 1).Our ref.: 240259EPWOP 21Sony Semiconductor Solutions Corporation According to this example, the resulting reliability value is set by the first condition met. Theresult may be reliable only if all checks (e.g., Order 1 to 6) pass.Fig.1 shows a table of the effects of the reliability value on the diagnostic indicators correlation, mean white intensity, mean black intensity, contrast and blur. The first column “order” indicates the order in which the different reliability values are determined (see also above). The second column “reliability value” indicates the different reliability values which may be determined as described above based on some of the diagnostic indicators and which indicate depending on the values the degradation of some of the measured diagnostic indicators. Columns 3 to 5 show the diagnostic indicators: correlation, mean white intensity, mean black intensity, contrast and blur and whether or not, based on the reliability value, the measurements of these diagnostic indicators are degraded, with some diagnostic indicators being unreliable andothers being under- or overestimated.The first row shows that the reliability value indicates “too low correlation”, which may be determined based on the correlation as described above. This indicates that all diagnostics but thecorrelation itself are unreliable and should not be used. Therefore, the mean white intensity,mean black intensity, contrast and blur diagnostic indicators are “Unreliable” and the correlation is "Reliable”. The second row shows that the reliability value indicates “too dark”, which may be determinedas described above. This indicates a degradation in the measurements of the diagnostic indicatorswith the correlation being underestimated, and the mean white intensity, mean black intensity,contrast and blur diagnostics indicators being unreliable.The third row shows that the reliability value indicates “too bright”, which may be determined asdescribed above. This indicates a degradation in the measurements of the diagnostic indicatorswith the correlation, the mean white intensity and the contrast being underestimated, the meanblack intensity being overestimated, and the blur being unreliable.The fourth row shows that the reliability value indicates “too blurred” which may be determinedas described above. This indicates a degradation in the measurements of the diagnostic indicatorswith the correlation, the mean white intensity and the contrast being underestimated, the meanblack intensity being overestimated and the blur being unreliable.The fifth row shows that the reliability values indicate “saturated”, which may be determined as described above. This indicates a degradation in the measurements of the diagnostic indicatorsOur ref.: 240259EPWOP 22Sony Semiconductor Solutions Corporationwith the correlation, the mean white intensity and the contrast being underestimated, the meanblack intensity being reliable and the blur being overestimated.The sixth row shows that the reliability values indicates “Unreliable”, which may be determined as described above. This indicates a degradation in the measurements of the diagnostic indicatorswith all diagnostic indicators being unreliable.The seventh row indicates that the reliability value is set to “Reliable”, as determined above (all checks passed, nothing is wrong). Therefore, all the diagnostic indicators are reliable. The order of the first column indicates in which order the determination of the reliability values is performed. Also, when all current parameters of the monitoring image are computed and compared with the reference parameters of the reference image, which may only happen if the reliability value is “Reliable” (see Order 7, Fig.1), the check may be fully passed.Otherwise, partial checks are possible. The monitoring image may be diagnosed to be notdegraded if partial checks are passed. If multiple marker regions are used, this may also be basedon the different marker regions.For example, if blur cannot be checked in one marker region, for example, because the markerregion is occluded, and the blur diagnostic indicator is discarded as unreliable (e.g., Order 1, Fig.1), but one or more other marker regions of the same monitoring image are not occluded, themonitoring image may be diagnosed to be not degraded. That is, because it may be determinedthat the failed blur check does not occur because of as loss of focus on the lens, as this wouldimpact all markers regions. In this way a partial check may lead to an image being diagnosed asnot degraded.In another example, part of the scene of a monitoring image may be too dark, but the part of thescene that is too dark may be outside of a monitored region of interest, (e.g., if some parts of themonitoring image are classified as monitoring regions of interest, for example, based onclassifying marker region(s) and / or the reference marker region(s) as monitoring region(s) ofinterest or as outside of monitoring region(s) of interest). In this case, it may be determined thatthe result “too dark” (e.g., Order 2, Fig. 1) is unproblematic, and the monitoring image may bediagnosed to be not degraded, for example if other checks are passed. In this way a partial checkmay lead to a monitoring image to be diagnosed as not degraded.In another example, when the correlation is tool low (e.g., reliability value is “Too lowcorrelation” in Order 1, Fig. 1), the correlation is reliable and it indicates (reliably) that the otherOur ref.: 240259EPWOP 23Sony Semiconductor Solutions Corporationdiagnostic indicators are unreliable (e.g., because of an occlusion). Otherwise, for example,when the correlation is underestimated (e.g., Order 2-5, Fig. 1) it may still be used, e.g., tocompare against a minimum threshold to determine whether there is an occlusion.Likewise, when the mean white intensity is underestimated (e.g., Order 3-5, Fig.1), it may stillbe used to determine whether there is enough light in the scene.Similarly, when the mean black intensity is overestimated (e.g., Order 4, Fig. 1), it may still beused to determine whether there is too much light in the scene.The same may apply for the contrast and the blur, which, when underestimated (e.g., Order 3-5,Fig. 1) may still be used to determine whether the contrast or the bur have not been too degraded.In the following other implementation examples of diagnostics are described, which may beperformed in any sequence:Occlusion DiagnosticThe correlation, as explained above, may be above a predefined threshold (e.g., a user threshold).In this case, the monitoring image may be diagnosed to not be degraded and / or the marker regionmay be diagnosed to still be visible in the same place and / or it may be determined that there is noocclusion.If the monitoring image is diagnosed to not be degraded, the camera is evaluated to be reliablefor monitoring.Alternatively, if it is determined that there is no occlusion, further diagnostics may be performedfor diagnosing if the monitoring image is degraded. For example, the following diagnostics may be performed. Object DiagnosticThe marker region(s) may be determined to be reliable with regard to the object diagnostic if,within tolerances, it has the same shape as the reference marker region and has not moved compared to the reference marker region. The tolerances may be based on predefined threshold(s). In this case, the monitoring image may be diagnosed to not be degraded and thus the camera is evaluated to be reliable for monitoring. Alternatively, further diagnostics may be performed for diagnosing if the monitoring image is degraded.Our ref.: 240259EPWOP 24Sony Semiconductor Solutions Corporation For example, the following diagnostics may be performed. YUV Image DiagnosticIt may be determined that the marker region has mean white and mean black intensities andcontrast within tolerances, for example, after comparison with mean white and mean blackintensities and contrast of the corresponding reference marker region, wherein tolerances may bebased on predefined threshold(s). In this case, the monitoring image may be diagnosed to not be degraded and thus the camera is evaluated to be reliable for monitoring. Alternatively, further diagnostics may be performed for diagnosing if the monitoring image is degraded. For example, the following or above diagnostics may be performed. Edge Sharpness DiagnosticThe marker region may be determined to have blur within tolerances, for example, aftercomparison of blur of the corresponding reference marker region, wherein tolerances may be based on predefined threshold(s). In this case, the monitoring image may be diagnosed to not be degraded and thus the camera is evaluated to be reliable for monitoring. Alternatively, further diagnostics may be performed for diagnosing if the monitoring image is degraded. For example, the following or above diagnostics may be performed. Ambient Light and Contrast DiagnosticThe marker region may be determined to have contrast and mean white intensities withintolerances, for example, after comparison with contrast and mean white intensities of thecorresponding reference marker region, wherein tolerances may be based on predefined threshold(s).In this case, the monitoring image may be diagnosed to not be degraded and thus the camera isevaluated to be reliable for monitoring. Alternatively, further diagnostics may be performed for diagnosing if the monitoring image is degraded. For example, the above diagnostics may be performed.Our ref.: 240259EPWOP 25Sony Semiconductor Solutions Corporation *** The abovementioned diagnostics may be performed on multiple marker regions if multiple marker regions are provided as explained above.The abovementioned diagnostics may be performed on all reliable marker regions. That is, if themarker region is determined to be unreliable in a later processing stage the previous measuresmay be discarded or marked as unreliable and subsequent diagnostics may be performed only onreliable markers. Also, to facilitate the abovementioned diagnostics, edge enhancement may not be performed during image processing. Based on the diagnostics described above the image quality of the monitoring image(s) can bediagnosed, and it can be evaluated whether the image(s) captured by the monitoring camera aregood enough to allow reliable monitoring, for example, such that a danger / accident or incident detection algorithm can operate reliably, e.g., always detects if personnel enters a predefined danger zone, or children leave a predefined safe zone etc..Furthermore, it has been recognized that the abovementioned diagnostics can run with a minimalfootprint making them suitable for widespread usage on the processors utilized in monitoringcameras (e.g., industrial cameras, security cameras etc.). In this way, as only a small subset ofpixels is used, the abovementioned implementations may require only a small CPU and memory usage. Also, any of the thresholds described above may be adapted to take into account the monitoringimages that occur naturally, e.g., in low light conditions, and any predefined thresholds orparameters of this specification may be defined by a user.Moreover, the abovementioned diagnostics may be combined with a registration process to beapplicable to multiple and moving cameras as described above.Furthermore, some embodiments, pertain to an electronic device including the circuitry with anyof the features of the circuitry described in this specification.Also, some embodiments, pertain to a method for diagnosing external interference influencing amonitoring image captured with a monitoring camera, comprising: obtaining reference intensities of a plurality of reference pixels of a reference marker region in a reference image captured with the monitoring camera, wherein the reference marker region is smaller than the reference image,Our ref.: 240259EPWOP 26Sony Semiconductor Solutions Corporation obtaining the monitoring image, determining current intensities of a plurality of sample pixels of a marker region in the monitoring image, wherein the marker region corresponds to the reference marker region, andwherein the plurality of sample pixels of the marker region corresponds to the plurality ofreference pixels of the reference marker region, determining a correlation between the determined current intensities and the reference intensities, and diagnosing, based on the determined correlation and a predefined threshold, the monitoring image for degradation due to an external interference for evaluating the reliability of themonitoring camera as a monitoring tool.The method may exhibit any feature described above with respect to the circuitry and / or anysuitable feature described below with respect to any one of the figures. The method may beperformed by the circuitry and / or by the electronic device.The methods as described herein are also implemented in some embodiments as a computerprogram causing a computer and / or a processor to perform the method, when being carried out on the computer and / or processor. In some embodiments, also a non-transitory computer- readable recording medium is provided that stores therein a computer program product, which, when executed by a processor, such as the processor described above, causes the methods described herein to be performed. Returning to Fig.2, an example of a reference marker region defined by virtual markers 1 and aselection of reference pixels based on the virtual marker layout is shown according to anembodiment.Reference image 14 is overlayed with a uniform grid 13 representing candidate virtual markerswhich can define reference marker regions and thereby can be used to sample reference pixelsfrom the reference image. The uniform grid 13 is based on an input pitch of 16. Each square inthe grid 13 is a Region Of Interest (ROI), from which a number of pixels can be sampled.In reference image 14 two virtual markers 1 are shown. On the left of Fig.2 an example of a virtual marker 1 defining the reference marker region, wherein reference pixels are sampled fromthe horizontal 10 and vertical 11 comprising a crosshair is shown. On the right of Fig. 2 a secondcrosshair virtual marker 1 defining a second reference marker region, wherein reference pixels are sampled from the crosshair and also the diagonals 12 is shown. In this way, virtual markers1 are selected from the grid 13. In reference image 14 virtual markers 1 define two referenceOur ref.: 240259EPWOP 27Sony Semiconductor Solutions Corporationmarker regions. Both virtual markers 1 have a pitch of 16, which results in 17 points on eitherhorizontal 10 or vertical 11 or diagonal 12 including one overlapping point in the middle. Thus,the left virtual marker 1 without diagonals 12 has 33 points (= 2*16+1), which results in 33reference pixels sampled for the left reference marker region. Correspondingly, the right virtualmarker 1 with diagonals 12 has 65 points (= 4*16+1) and therefore 65 reference pixels aresampled from the right reference marker region. In this way, the grid 13 is used to sample thereference image. (The squares of Fig.1 surrounding the virtual markers 1 are for illustration purposes only.) That is, reference pixels are selected based on points on the lines 10, 12, 13 of each virtualmarker 1. That is, the points on the crosshair 10, 11 on the left or the crosshair 10, 11 anddiagonals 12 on the right correspond to candidate reference pixels for the two respective reference marker regions, from which the plurality of reference pixels are selected.The reference image 14 may be a static image.The sampling of reference pixels may be carried out as part of the initialization (see Figs, 7 and8). Concerning the monitoring image(s) (e.g., 22, Figs.4,5), each monitoring image may be sampled using the same grid 13. The joint statistics of the reference image and the monitoring image maythen be obtained. For example, a noise-adjusted Pearson correlation as described above, forexample with ^^^^ = 0, ^^^^ = 3, may be computed and compared to a predefined threshold,for example, ^^^^^^^= 0.913. Also, the contrast ratio as described above may be determined andcompared to a predefined ^^^^^^^^^^^^^^^^ , for example ^^^^^^^^^^^^^^^^ = 0.611, and theblur ratio may be determined and compared to a predefined ^^^^^^^^^^^^ , for example^^^^^^^^^^^^ = 2.38.Fig. 3 shows an example of a real marker 1 defining the reference marker region and the markerregion. The real marker 1 can be placed anywhere in the scene from where the monitoringcamera can capture it for initialization (see Figs. 7 and 8).Real marker 1 includes four concentric circles 2-5 of decreasing radius with a common center where the colors alternate between black and white. Starting from the center, the innermostcenter circle 5 is white, the second circle 2 is black, the third circle 3 is white and the outermostfourth circle 4 is black again.Center ring 5 has a diameter of 3X, where X is 3 pixels of the reference image, so 9 pixels intotal. The second circle 2 has a size of 2X pixels so the added diameter is 7X. The third circle 4Our ref.: 240259EPWOP 28Sony Semiconductor Solutions Corporation has a diameter of X, so that the added diameter is 9X and the fourth circle also has a diameter of X. In total, the diameter D of the real marker is 11X, which amounts to 33 pixels of the reference image. The ratio between black and white parts of the real marker 1 is 1 / 3. The third and fourth circles 3,4 are outer circles and have a high frequency, because the black and white variation occurs in fewer pixels. The center ring 5 is thicker than all the other rings and therefore has a low frequency in comparison. Concerning diagnostics, the mean white intensity may be based on pixels falling in the centerring 5. The mean black intensity may be based on pixels falling in the second circle 2. Thecontrast may be calculated based on the contrast of the center circle 1 and the second circle 2.The blur may be based on the number of pixels needed to switch from white to black transition between center circle 1 and second circle 2.Similarly, to the virtual markers 1 of Fig. 2 lines, e.g. a vertical, horizontal or diagonal line maybe used to select the reference pixels, similar to the vertical 10, the horizontal 11 and thediagonals 12 illustrated in Fig. 2. For example, the reference and sample pixels may be selectedfrom the respective reference marker region and marker region in four directions from the center crossing all four circles 2-5. In this way, pixels from all circles may be selected and transitions,safe areas and low and high frequency areas may be determined corresponding to the selectedpixels.Fig. 4 illustrates an embodiment for diagnosing external interference influencing a monitoringimage captured with a monitoring camera. Reference intensities 21a, which are the intensities of a plurality of reference pixels selected from a reference marker region in a reference image captured by the monitoring camera, are obtained and provided to the correlation 25a. For example, reference intensities 21a are obtainedfrom a storage (not shown), which may be stored in the storage after initialization (see Figs. 7and 8).Monitoring image 22 captured by the monitoring camera is obtained. Based on monitoring image22 current intensities 24a are determined in a determination process 23. During the determination23 a marker region in monitoring image 22 which corresponds to the reference marker region is determined and a plurality of sample pixels which correspond to the plurality of reference pixels are selected. Current intensities 24a of the plurality of sample pixels are then determined andprovided to correlation 25a.Our ref.: 240259EPWOP 29Sony Semiconductor Solutions Corporation During the correlation process 25a a correlation is calculated between reference intensities 21a and current intensities 24a and the correlation coefficient is determined. Furthermore, predefinedthreshold 27 is obtained and based on predefined threshold 27 diagnostics (see e.g., 25, Fig. 5) ofthe monitoring image 22 are performed, which results in the diagnosis 26. For example, if thedetermined correlation coefficient is larger than (or equal to) the predefined threshold 27, which,for example, is a predefined threshold of 0.913, then the monitoring image 22 is diagnosed to benot degraded in diagnosis 26 and therefore, the monitoring camera is evaluated to be performing reliably for monitoring. On the other hand, it the determined correlation coefficient is lower than the predefined threshold than the monitoring image 22 is diagnosed to be degraded in diagnosis 26 and therefore, the monitoring camera is evaluated to be performing unreliably for monitoring.Fig. 5 illustrates an embodiment for diagnosing external interference influencing a monitoringimage captured with a monitoring camera using multiple diagnostics. Reference parameters 21, including mean white intensity 21b, mean black intensity 21c, blur 21e, contrast 21d and reference intensities 21a are obtained. Reference intensities 21a are the intensities of a plurality of reference pixels selected from a reference marker region in a reference image captured by the monitoring camera, as explained in Fig.4. Mean white intensity 21b is the average intensity of reference pixels of the predefined white parts of the reference marker region, for example the reference pixels falling into the center circle 1 of real marker 1 of Fig.3.Mean black intensity 21c is the average intensity of those reference pixels of the predefinedblack parts of the reference marker region, for example the reference pixels falling into the second circle 2 of real marker 1 of Fig.3. Blur 21e is calculated based on ^^ ^^^^^^^^^=^^^^^,^^^, where ^^^^= is the standard deviation of the reference intensities 21a,^^,^^^ is the reference intensity 21a of each reference pixel of the plurality of reference pixels,^^^^= is the mean reference intensity 21a of the plurality of reference pixels,Our ref.: 240259EPWOP 30Sony Semiconductor Solutions Corporation ^^^^,^^^is the maximum difference in reference intensities 21a between adjacent pixels of the plurality of reference pixels. Contrast 21d is calculated based on defined above for blur 21e. Alternatively, if real marker 1 of Fig.3 is used for defining the reference marker region and the marker region, the contrast between center circle 5 and second circle 2 may be calculated as contrast 21d based on the Michelson contrast: ^^^^^^ is the intensity of reference pixels of the white part of the reference marker region, e.g., ofcenter circle 5 or circles 5 and 3 of Fig.3, and ^^^^^^is the intensity of reference pixels of the black part of the reference marker region, e.g., of center circle 5 of Fig.3 or circles 5 and 3 of Fig.3.Reference parameters 21 are obtained from a storage (not shown), which may be stored in thestorage after calculation of the reference parameters 21 during an initialization (see Figs. 7 and8), and provided to diagnostics 25. Alternatively, only some reference parameters, for example,only the reference intensities 21a are obtained from storage and the other reference parameters 21b-e may be determined based on the obtained reference intensities 21a. Furthermore, reference parameters 21b-e may be determined based on reference intensities 21a obtained from storage and other parameters obtained from storage, which may be determined during the initialization.Such other parameters may, for example, be information on different areas of the referencemarker region, e.g., black part, white part, high frequency area, low frequency area, transitionlocation, safe location (i.e., far from transition locations according to a predefined threshold) etc.Monitoring image 22 captured by the monitoring camera is obtained, for example, from storage. Based on monitoring image 22 current parameters 24 are determined in a determination process 23. During the determination 23 a marker region in monitoring image 22 which corresponds to the reference marker region is determined and a plurality of sample pixels which correspond to the plurality of reference pixels are selected. Current intensities 24a of the plurality of sample pixels as well as mean white intensity 24b, mean black intensity 25c, blur 24e and contrast 24d are determined and provided to diagnostics 25.Our ref.: 240259EPWOP 31Sony Semiconductor Solutions CorporationMean white intensity 24b is the average intensity of those sample pixels of the predefined whiteparts of the marker region, for example the sample pixels falling into the center circle 1 of realmarker 1 of Fig.3. Mean black intensity 24c is the average intensity of those reference pixels of the predefined black parts of the marker region, for example the sample pixels falling into the second circle 2 of real marker 1 of Fig.3. Blur 21e is calculated based on ^^ ^^^^^^^=^^^^^^^,^^^, where ^^^^,^^^ is the maximum difference in current intensities 24a between adjacent pixels of theplurality of sample pixels. Contrast 24d is calculated based on ^^^^^^^^^^^= ^^^^, with the parameters as defined above for blur 24e. Alternatively, if real marker 1 of Fig.3 is used for defining the reference marker region and the marker region, the contrast between white and black parts of the marker region, e.g., center circle 5 and second circle 2 of Fig.3, may be calculated as contrast 21d based on the Michelson contrast: ^^^^^^ is the intensity of sample pixels of the white part of the marker region, e.g., of center circle5 or circles 5 and 3 of Fig.3, and ^^^^^^is the intensity of sample pixels of the black part of the marker region, e.g., of center circle 5 of Fig.3 or circles 5 and 3 of Fig.3. For the diagnostics 25, correlation 25a, as described in Fig.4, mean white intensity comparison25b, mean black intensity comparison 25c, blur ratio 25e, and contrast ratio 25d are determined.Mean white intensity comparison 25b calculation includes a comparison between mean whiteintensity 24b and mean white intensity 21b. If the mean white intensity 24b falls within predefined tolerances, e.g. a predefined threshold, of the mean white intensity 21b, it isOur ref.: 240259EPWOP 32Sony Semiconductor Solutions Corporation determined that the marker region of the monitoring image passes the mean white intensitycheck. Otherwise, it fails the mean black intensity check, for example because mean blackintensity 24c is overestimated compared to mean black intensity 21c or because mean black intensity 24c is underestimated compared to mean black intensity 21c. Mean black intensity comparison 25c calculation includes a comparison between mean blackintensity 24c and mean black intensity 21c. If the mean white intensity 24b falls withinpredefined tolerances, e.g., a predefined threshold, of the mean black intensity 21c, it isdetermined that the marker region of the monitoring image passes the mean black intensitycheck. Otherwise, it fails the mean black intensity check, for example because mean blackintensity 24c is overestimated compared to mean black intensity 21c or because mean black intensity 24c is underestimated compared to mean black intensity 21c. Blur ratio 25e is determined based on: ^^^^^^^^^^^^^^If blur ratio 25e falls within predefined tolerances, e.g., is smaller or equal to a predefinedthreshold ^^^^^^^^^^^^, it is determined that the marker region of the monitoring image passesthe blur check. Otherwise, it fails the blur check, for example if it is overestimated or underestimate. The ^^^^^^^^^^^^may for example be 2.38. Contrast ratio 25d is determined based on: ^^^^^^^^^^^^^^^^^^^^^^If the Michelson contrast is used, then instead or additionally a comparison of Michelsen contrasts between the marker region and the reference marker region may be used.If contrast ratio 25d falls within predefined tolerances, e.g., is larger or equal to predefinedthreshold ^^^^^^^^^^^^^^^^, it is determined that the marker region of the monitoring imagepasses the contrast check. Otherwise, it fails the contrast check, for example if it is overestimated or underestimated. The ^^^^^^^^^^^^^^^^may for example be 0.611.Correlation 25a is calculated as described in Fig. 4. If the determined correlation coefficient fallswithin predefined tolerances, e.g., larger or equal to predefined threshold 27 of Fig. 4, it isdetermined that the marker region of the monitoring image passes the correlation check. Otherwise, it fails the correlation check, for example if it is overestimated or underestimate. The predefined threshold 27 may for example be 0.913.Our ref.: 240259EPWOP 33Sony Semiconductor Solutions Corporation Diagnostics 25 may be performed in any order and based on diagnostics 25, further diagnostics,for example the determination of the reliability value as illustrated in Fig. 1 and described abovemay be performed. Based on diagnostics 25, which may include further diagnostics and the reliability value determination of Fig.1, a diagnosis 26 results. For example, if all checks are fully passed, themonitoring image 22 may be diagnosed positively to be not degraded in diagnosis 26 andtherefore, the monitoring camera is evaluated to be performing reliably for monitoring. On the other hand, if all checks are failed the monitoring image 22 may be diagnosed negatively to be degraded in diagnosis 26 and therefore, the monitoring camera is evaluated to be performing unreliably for monitoring. Additionally, if multiple reference marker regions and therefore multiple marker regions areused, then the above-described procedures (determination 23, diagnostics 25, obtaining referenceparameters 21) may be conducted for each marker region.In this case, as diagnostics 25 are marker region dependent, the resulting diagnosis 26 may be positive, i.e., the image being diagnosed as not degraded, if a predefined percentage of all markerregions are positively diagnosed, e.g., if over 50% of the marker regions pass all checks fully.Also, multiple cameras may be used to capture multiple corresponding monitoring images. For example, a stereoscopic left and right camera may be used to capture the same scene in corresponding monitoring images 22, in such a way that all of the multiple monitoring images 22 include the marker region. In this case, determination of current parameters 23 for each of the multiple corresponding monitoring images 22 may be performed. Also, diagnostics 25 may be performed between the current parameters 24 of each of the multiple monitoring images 22 and the reference parameters 21, but also between the current parameters 24 of the multiple monitoring images. For example, the mean white intensity 24b of a right monitoring image captured by the right stereoscopic camera may be compared to the mean white intensity 24b of the left monitoring image 22 captured by the left stereoscopic camera. Similarly, also multiple cameras may be used to capture multiple corresponding reference images. Then, also reference parameters for each of the reference images may be obtained and diagnostics 25 may be based on the reference parameters for each of the reference images.Fig. 6 illustrates an embodiment of a method for diagnosing external interference influencing amonitoring image captured with a monitoring camera.Our ref.: 240259EPWOP 34Sony Semiconductor Solutions Corporation In step 41 reference intensities (21a, Fig.5) of a plurality of reference pixels of a reference marker region in a reference image captured with a monitoring camera are obtained.In step 42 a monitoring image (22, Figs. 4,5), the monitoring image being captured by themonitoring camera, is obtained. In step 43 the current intensities (24a, Fig.5) of a plurality of sample pixels of a marker region inthe monitoring image (22, Figs. 4,5) are determined, wherein the marker region corresponds tothe reference marker region, and wherein the plurality of sample pixels of the marker region corresponds to the plurality of reference pixels of the reference marker region. In step 44 the correlation (25a, Figs.4,55) between the determined current intensities (24b, Figs. 4,5) and the reference intensities (21a, Figs.4,5) are obtained. In step 45, based on the determined correlation (25a, Figs.4,5) and a predefined threshold (27,Fig. 4) the monitoring image (22, Figs. 4,5) is diagnosed (25, 26, Fig. 4,5) for degradation due toan external interference for evaluating the reliability of the monitoring camera as a monitoring tool. The correlation of step 44 may be based on a noise-adjusted Pearson correlation, wherein thenoise-adjusted Pearson correlation coefficient ^^^^ is calculated according to: n is the number of reference pixels as well as the number of sample pixels (e.g., n = 10, then 10sample pixels and 10 reference pixels are compared)^^,^^^ is the reference intensity (21a, Figs, 4,5) of each reference pixel of the plurality ofreference pixels,^^,^^^is the current intensity (24a, Figs, 4,5) of each sample pixel of the plurality of sample pixels, ^ ^^^^= ^∑^ ^^^,^^^is the mean reference intensity (21a, Figs, 4,5) of the plurality of reference pixels, ^ ^^^^= ^∑^ ^^^,^^^is the mean current intensity (24a, Figs, 4,5) of the plurality of sample pixels,Our ref.: 240259EPWOP 35Sony Semiconductor Solutions Corporation ^ ^^^^= ^ is the standard deviation of the reference intensities (21a, Figs,4,5), ^^^^ the standard deviation of the current intensities (21a, Figs, 4,5),and^^^^ and ^^^^ are predefined correction coefficients.Fig. 7 illustrates an example of an initialization process for determining reference parametersaccording to an embodiment. In step 51 a reference image is captured, for example with the monitoring camera. In step 52 the reference marker region in the captured reference image is determined, for example as defined by the real marker 1 of Fig.3 or as defined by any of the virtual markers 1 of Fig.1.In step 53 the reference pixels in the reference marker region are selected, for example, asillustrated and described with regard to Figs. 1 and 3In step 54, reference parameters (e.g.21 of Fig.5) including the reference intensities (21a, Figs, 4,5) of the selected reference pixels are determined. Furthermore, other reference parameters, such as the reference parameters 21b-e of Fig.5, may be determined in step 54. Alternatively, or additionally also other parameters such as any one or more of the group of black part(s), white part(s), transition location(s), safe area(s), high frequency area(s), low frequency area(s) of the reference marker region may be determined. For example, transition locations may be based on transitions between colors (e.g., white toblack) or transitions between high to low frequency areas.Transition locations may be determined based on stepwise detection of pixel intensities on ahorizontal vertical or diagonal line, wherein the crossing threshold, when the transition from onecolor to another color (e.g., black to white) is detected in a reference pixel, may be 50% of therange of intensities, for example if the intensities range from 10 to 46, the 50% threshold is at 28. For example, delimitation of safe areas may be based on the transition locations. After step 54, all determined reference parameters and / or other parameters, from which reference parameters may be determined, may be stored in a storage.Our ref.: 240259EPWOP 36Sony Semiconductor Solutions CorporationFor example, transition locations, reference pixel locations for each safe area and referenceintensities and / or reference intensities for low frequency white and black parts may be stored in storage. Furthermore, image processing techniques may be used after step 54, but before storage of the reference intensities. For example, gamma compression, e.g., a gamma compression of 0.5, may be applied. Additionally, also diagnostics may be performed with regard to the reference marker region.For example, Fig. 8 illustrates an example of a contrast check of a reference marker regionaccording to an embodiment.Steps 51 to 54 proceed as described with regard to Fig. 7. Alternatively, to steps 51 to 54, thereference intensities may be obtained from storage.In step 55 the reliability of contrast of the reference marker region is determined based on a Pearson correlation between the reference intensities and obtained current intensities 24a as described in Figs.4 and 5. The Pearson correlation may be based on the following: n is the number of reference pixels as well as the number of sample pixels,is the reference intensity (21a, Figs, 4,5) of each reference pixel of the pluralityreference pixels,^^,^^^is the current intensity (24a, Figs, 4,5) of each sample pixel of the plurality of sample pixels, ^^^^ is the mean reference intensity (21a, Figs, 4,5) of the plurality of reference pixels, ^ ^^^^= ^∑^ ^^^,^^^is the mean current intensity (24a, Figs, 4,5) of the plurality of sample pixels, ^ = is the standard deviation of the reference intensities (21a, 4,5),Our ref.: 240259EPWOP 37Sony Semiconductor Solutions Corporation ^ ^^^^= ^^∑^^ ^^^,^^^ − ^^^^^^is the standard deviation of the current intensities (21a, Figs, 4,5).The Pearson correlation coefficient computed according to the above formula ^^^^^^^ as wellas a predefined threshold may be used to determine whether the contrast is reliable, e.g., so that the reference marker region can be used as intended in later computations (see Figs.4-6).For example, if ^^^^^^^ is higher than a predefined threshold, e.g., 0.9, then the referencemarker region is determined to have a reliable contrast.If, on the other hand, the ^^^^^^^ is, for example, lower or equal to the predefined threshold,e.g., 0.9, then the reference marker region is determined to have an unreliable contrast.Alternatively, if ^^^^^^^ is higher or equal to a predefined threshold, e.g., 0.9, then thereference marker region may be determined to have a reliable contrast and if the ^^^^^^^ islower than the predefined threshold, e.g., 0.9, then the reference marker region may be determined to have an unreliable contrast. In that case, the reference marker region is discarded, and another reference marker region is selected during step 52. Also, other diagnostics may be performed for the reference marker region. For example, if the real marker 1 of Fig.3 is used, it may be determined whether the center circle 2 has an intensity(e.g., after gamma compression of 0.5) higher than or equal to a predefined threshold, forexample, the predefined threshold may be 40. Also, the noise of the reference image may be checked. For example, it may be determined whether the reference image is lower than 1 LSB after gamma compression of 0.5, which typically corresponds to an image with a gain of 1. If that is the case, then the reference image passes the noise check. If the reference marker region fails the diagnostics, then the reference marker region may be discarded and another reference marker region may be selected in step 52.Fig. 9 shows a block diagram depicting an embodiment of an electronic device for diagnosingexternal interference influencing a monitoring image captured with a monitoring camera.The electronic device 100 comprises a CPU 101 as processor. The electronic device 100 furthercomprises sensors(s) 106 connected to the processor 101. The processor 101 may for exampleimplement the processes described with regard to Figs. 4 to 8.Our ref.: 240259EPWOP 38Sony Semiconductor Solutions Corporation The sensors 106, may for example include a microphone. The microphone may be configured to receive any kind of audio signal. The sensors 106, may for example include camera(s). The camera may be any type of monitoring camera, such as an RGB camera, a stereoscopic camera, a depth sensing camera, an event-basedcamera or the like. It may be a movable camera or a static camera.The electronic device 100 further comprises loudspeaker(s) 108, such as headphones, e.g., on-ear, in-ear, over-ear, wireless headphones and the like, or may consist of one or more loudspeakers that are distributed over a predefined space and is configured to render any kind of audio. The electronic device 100 further comprise a user interface 107 that is connected to the processor 101. This user interface 107 acts as a man-machine interface and enables a dialogue between a user and the device. For example, a user may make configurations to the system using this user interface 107, such as, for defining any of the abovementioned thresholds (e.g.27, Fig.4).The electronic device 100 further comprises a Bluetooth interface 104 and a WLAN interface105. These units 104 and 105 act as I / O interfaces for data communication with external devices.For example, additional loudspeakers, microphones, and cameras (e.g., the monitoring camera)with WLAN or Bluetooth connection may be coupled to the processor 100 via these interfaces104, 105. The electronic device 100 further comprises a data storage 102 and a data memory 103 (here a RAM). The data memory 103 is arranged to temporarily store or cache data or computer instructions for processing by the processor 101. The data storage 102 is arranged as a long-termstorage, e.g., for reference parameters or other parameters as described in relation to Figs. 4 to 8.It should be noted that the description above is only an example configuration. Alternative configurations may be implemented with additional or other sensors, storage devices, interfaces, or the like. The electronic device 100 may be a stationary electronic device, such as the monitoring camera, a laptop computer, a personal computer, or a mobile device of any other kind of portable orwearable device, for example, a mobile monitoring camera, a smartphone, a tablet computer,smart glasses, head mounted displays (HMDs), earphones or other types of smart wearable devices, or the like. It should be recognized that the embodiments describe methods with an exemplary ordering ofmethod steps. The specific ordering of method steps is however given for illustrative purposesOur ref.: 240259EPWOP 39Sony Semiconductor Solutions Corporationonly and should not be construed as binding. For example, the ordering of 41 and 42 in theembodiment of Fig. 6 may be exchanged. Also, the ordering of 41 and 43 in the embodiment ofFig. 6 may be exchanged. Other changes of the ordering of method steps may be apparent to theskilled person.Please note that the division into determination of current parameters 23 and correlation 25a inFig. 4 or diagnostics 25 in Fig. 5 as well as the division for example within the diagnostics 25 isonly made for illustration purposes and that the present disclosure is not limited to any specificdivision of functions in specific units. For instance, the determination of current parameters 23 ordiagnostics 25 could be implemented by a respective programmed processor, fieldprogrammable gate array (FPGA) and the like.A method for controlling an electronic device (e.g. 100 of Fig. 9), is described under reference ofFig.6. The method can also be implemented as a computer program causing a computer and / or aprocessor, such as processor 101 discussed above, to perform the method, when being carried outon the computer and / or processor. In some embodiments, also a non-transitory computer- readable recording medium is provided that stores therein a computer program product, which, when executed by a processor, such as the processor described above, causes the method described to be performed. All units and entities described in this specification and claimed in the appended claims can, if not stated otherwise, be implemented as integrated circuit logic, for example on a chip, and functionality provided by such units and entities can, if not stated otherwise, be implemented by software. In so far as the embodiments of the disclosure described above are implemented, at least in part, using software-controlled data processing apparatus, it will be appreciated that a computer program providing such software control and a transmission, storage or other medium by whichsuch a computer program is provided are envisaged as aspects of the present disclosure.Note that the present technology can also be configured as described below.[1] Circuitry for diagnosing external interference influencing a monitoring image (22)captured with a monitoring camera (106), configured to: obtain reference intensities (21a) of a plurality of reference pixels of a reference marker region in a reference image captured with the monitoring camera (106), wherein the reference marker region is smaller than the reference image, obtain the monitoring image (22),Our ref.: 240259EPWOP 40Sony Semiconductor Solutions Corporation determine current intensities (24a) of a plurality of sample pixels of a marker region in the monitoring image (22), wherein the marker region corresponds to the reference marker region, and wherein the plurality of sample pixels of the marker region corresponds to the plurality of reference pixels of the reference marker region, determine a correlation between the determined current intensities (24a) and the reference intensities (21a), and diagnose, based on the determined correlation and a predefined threshold (27, ^^^^^^^), the monitoring image (22) for degradation due to an external interference for evaluating the reliability of the monitoring camera (106) as a monitoring tool.[2] The circuitry of [1], wherein the correlation is a Pearson correlation.[3] The circuitry of [1] or [2], wherein determining the correlation includes determining anoise-adjusted Pearson correlation coefficient ^^^^ according to the following formula:wherein nis the number of reference pixels as well as the number of sample pixels,^^,^^^ is the reference intensity (21a) of each reference pixel of the plurality ofreference pixels, wherein^^,^^^is the current intensity (24a) of each sample pixel of the plurality of sample pixels, wherein ^ ^^^^= ^∑^ ^^^,^^^is the mean reference intensity (21a) of the plurality of reference pixels, wherein ^ = is the mean current intensity (24a) of the plurality of pixels, wherein ^ = is the standard deviation of the reference (21a), whereinOur ref.: 240259EPWOP 41Sony Semiconductor Solutions Corporation ^ ^^^^= ^ is the standard deviation of the current intensities (24a), wherein ^^^^ and ^^^^ are predefined correction coefficients.[4] The circuitry of [3], wherein the monitoring image (22) is diagnosed as degraded and themonitoring camera (106) as unreliable for monitoring if the noise-adjusted Pearson correlationcoefficient ^^^^ is lower than the predefined threshold (27, ^^^^^^^).[5] The circuitry of any one of [1] to [4], wherein the reference intensities (21a) are based onthe luminance values of the plurality of reference pixels and the current intensities (24a) are based on the luminance values of the plurality of sample pixels.[6] The circuitry of any one of [1] to [5], wherein the reference marker region has a reliablecontrast, wherein reliability of the contrast is based on a Pearson correlation coefficient of a Pearson correlation between the reference intensities (21a) and the current intensities (24a) being above a predefined contrast threshold.[7] The circuitry of [6], wherein the contrast threshold is at least 0.9.[8] The circuitry of [6] or [7], wherein the contrast is the greyscale contrast or the colorcontrast.[9] The circuitry of any one of [1] to [8], wherein the circuitry is further configured todetermine a mean white intensity (24b) of the plurality of sample pixels, and / or a mean black intensity (24c) of the plurality of sample pixels, and / or a contrast (24d) of the plurality of sample pixels, and / or a blur (24e) of the plurality of sample pixels, and wherein diagnosing the monitoring image is further based on the mean white intensity (24b) of the plurality of sample pixels, and / or the mean black intensity (24c) of the plurality of sample pixels, and / or the mean black intensity (24d) of the plurality of sample pixels, and / or the blur (24e) of the plurality of sample pixels.

[0010] The circuitry of [9], wherein the circuitry is further configured to obtainOur ref.: 240259EPWOP 42Sony Semiconductor Solutions Corporation a mean white intensity (21b) of the plurality of reference pixels, and / or a mean black intensity (21c) of the plurality of reference pixels, and / or a contrast (21d) of the plurality of reference pixels, and / or a blur (21e) of the plurality of reference pixels, wherein the circuitry is further configured to determine a mean white intensity comparison (25b) between the mean white intensity (24b) of the plurality of sample pixels and the mean white intensity (21b) of the plurality of reference pixels, and / or a mean black intensity comparison (25c) between the mean black intensity (24c) of the plurality of sample pixels and the mean white intensity (21c) of the plurality of reference pixels, and / or a contrast ratio (25d) between the contrast (24d) of the plurality of sample pixels and the contrast (21d) of the plurality of reference pixels, and / or a blur ratio (25e) between the blur (24e) of the plurality of sample pixels and the blur (21e) of the plurality of reference pixels, and wherein diagnosing the monitoring image (22) is further based on the mean white intensity ratio (25b) and / or the mean black intensity ratio (25c) and / or the contrast ratio (25d), and / or the blur ratio (25e).

[0011] The circuitry of

[0010] ,^^^^^^^^ wherein the contrast ratio (25d) is determined according to^^^^^^^^^^^^^^, ^^^^ wherein the blur ratio is determined to wherein ^ ^^^^^^^^^^^=^^^^^^^, wherein the mean current intensity (24a) of the plurality of sample pixels, wherein ^^^^= is the mean reference intensity (21a) of the plurality of reference pixels, whereinOur ref.: 240259EPWOP 43Sony Semiconductor Solutions Corporation ^ ^^^^= ^^∑^^ ^^^,^^^ − ^^^^^^is the standard deviation of the current intensities (24a), wherein ^^^^ is the standard deviation of the reference intensities (21a), wherein ^^ ^^^^^^^^^^= ^^^^,^^^, wherein ^^^^,^^^is the maximum difference in current intensities (24a) between adjacent pixels of the plurality of sample pixels, wherein ^^ ^^^^^^^^^^= ^^^^,^^^, wherein ^^^^,^^^is the maximum difference in reference intensities (21a) between adjacent pixels of the plurality of reference pixels.

[0012] The circuitry of any one of [1] to

[0011] , wherein the reference marker region and themarker region are defined based on a virtual marker (1).

[0013] A method for diagnosing external interference influencing a monitoring image (22)captured with a monitoring camera (106), comprising: obtaining reference intensities (21a) of a plurality of reference pixels of a reference marker region in a reference image captured with the monitoring camera (106), wherein the reference marker region is smaller than the reference image, obtaining the monitoring image (22), determining current intensities (24a) of a plurality of sample pixels of a marker region in the monitoring image (22), wherein the marker region corresponds to the reference marker region, and wherein the plurality of sample pixels of the marker region corresponds to the plurality of reference pixels of the reference marker region, determining a correlation between the determined current intensities (24a) and the reference intensities (21a), and diagnosing, based on the determined correlation and a predefined threshold (27), the monitoring image (22) for degradation due to an external interference for evaluating the reliability of the monitoring camera (106) as a monitoring tool.

[0014] The method of

[0013] , wherein the correlation is a Pearson correlation.Our ref.: 240259EPWOP 44Sony Semiconductor Solutions Corporation

[0015] The method of

[0013] or

[0014] , wherein determining the correlation includes determining anoise-adjusted Pearson correlation coefficient ^^^^ according to the following formula:wherein nis the number of reference pixels as well as the number of sample pixels,^^,^^^ is the reference intensity (21a) of each reference pixel of the plurality ofreference pixels, wherein^^,^^^is the current intensity (24a) of each sample pixel of the plurality of sample pixels, wherein is the mean reference intensity (21a) of the plurality of reference pixels, wherein ^ = is the mean current intensity (24a) of the plurality of pixels, wherein ^^^^= the standard deviation of the reference intensities (21a), wherein is the standard deviation of the current (24a), wherein ^^^^ and ^^^^ are predefined correction coefficients.

[0016] The method of

[0015] , wherein the monitoring image (22) is diagnosed as degraded and themonitoring camera (106) as unreliable for monitoring if the noise-adjusted Pearson correlationcoefficient ^^^^ is lower than the predefined threshold (27, ^^^^^^^).

[0017] The method of any one of

[0013] to

[0016] , wherein the reference intensities (21a) are basedon the luminance values of the plurality of reference pixels and the current intensities (24a) are based on the luminance values of the plurality of sample pixels.Our ref.: 240259EPWOP 45Sony Semiconductor Solutions Corporation

[0018] The method of any one of

[0013] to

[0017] , wherein the reference marker region has a reliablecontrast, wherein reliability of the contrast is based on a Pearson correlation coefficient of a Pearson correlation between the reference intensities (21a) and the current intensities (24a) being above a predefined contrast threshold.

[0019] The method of

[0018] , wherein the contrast threshold is at least 0.9.

[0020] The method of any one of

[0018] to

[0019] , wherein the contrast is the greyscale contrast orthe color contrast.

[0021] The method of any one of

[0013] to

[0020] , wherein the method further comprises determininga mean white intensity (24b) of the plurality of sample pixels, and / or a mean black intensity (24c) of the plurality of sample pixels, and / or a contrast (24d) of the plurality of sample pixels, and / or a blur (24e) of the plurality of sample pixels, and wherein diagnosing the monitoring image is further based on the mean white intensity (24b) of the plurality of sample pixels, and / or the mean black intensity (24c) of the plurality of sample pixels, and / or the mean black intensity (24d) of the plurality of sample pixels, and / or the blur (24e) of the plurality of sample pixels.

[0022] The method of any one of [1] to

[0021] , wherein the circuitry further comprises obtaininga mean white intensity (21b) of the plurality of reference pixels, and / or a mean black intensity (21c) of the plurality of reference pixels, and / or a contrast (21d) of the plurality of reference pixels, and / or a blur (21e) of the plurality of reference pixels, wherein the method further comprises determining a mean white intensity comparison (25b) between the mean white intensity (24b) of the plurality of sample pixels and the mean white intensity (21b) of the plurality of reference pixels, and / or a mean black intensity comparison (25c) between the mean black intensity (24c) of the plurality of sample pixels and the mean white intensity (21c) of the plurality of reference pixels, and / orOur ref.: 240259EPWOP 46Sony Semiconductor Solutions Corporation a contrast ratio (25d) between the contrast (24d) of the plurality of sample pixels and the contrast (21d) of the plurality of reference pixels, and / or a blur ratio (25e) between the blur (24e) of the plurality of sample pixels and the blur (21e) of the plurality of reference pixels, and wherein diagnosing the monitoring image (22) is further based on the mean white intensity ratio (25b) and / or the mean black intensity ratio (25c) and / or the contrast ratio (25d), and / or the blur ratio (25e).

[0023] The method of

[0022] ,^^^^^^^^ wherein the contrast ratio (25d) is determined according to^^^^^^^^^^^^^^, wherein the blur ratio (25e) is determined according to wherein the mean current intensity (24a) of the plurality of sample pixels, wherein ^ = is the mean reference intensity (21a) of the plurality reference pixels, wherein is the standard deviation of the current (24a), wherein ^^^^ is the standard deviation of the reference intensities (21a), wherein ^^ ^^^^^^^^^^= ^^^^,^^^, wherein ^^^^,^^^is the maximum difference in current intensities (24a) between adjacent pixels of the plurality of sample pixels, wherein ^^ ^^^^^^^=^^^^^^^,^^^, whereinOur ref.: 240259EPWOP 47Sony Semiconductor Solutions Corporation ^^^^,^^^is the maximum difference in reference intensities (21a) between adjacent pixels of the plurality of reference pixels.

[0024] The method of any one of

[0013] to

[0023] , wherein the reference marker region and themarker region are defined based on a virtual marker (1).

[0025] A computer program comprising program code causing a computer to perform themethod according to anyone of

[0013] to

[0024] , when being carried out on a computer.

[0026] A non-transitory computer-readable recording medium that stores therein a computerprogram product, which, when executed by a processor, causes the method according to anyoneof

[0013] to

[0024] to be performed.

Claims

Our ref.: 240259EPWOP 1Sony Semiconductor Solutions Corporation CLAIMS1. Circuitry for diagnosing external interference influencing a monitoring image capturedwith a monitoring camera, configured to: obtain reference intensities of a plurality of reference pixels of a reference markerregion in a reference image captured with the monitoring camera, wherein the reference markerregion is smaller than the reference image, obtain the monitoring image,determine current intensities of a plurality of sample pixels of a marker region inthe monitoring image, wherein the marker region corresponds to the reference marker region,and wherein the plurality of sample pixels of the marker region corresponds to the plurality ofreference pixels of the reference marker region, determine a correlation between the determined current intensities and thereference intensities, and diagnose, based on the determined correlation and a predefined threshold, themonitoring image for degradation due to an external interference for evaluating the reliability ofthe monitoring camera as a monitoring tool.

2. The circuitry of claim 1, wherein determining the correlation includes determining anoise-adjusted Pearson correlation coefficient ^^^^ according to the following formula:wherein^ ^^^^^^,^^^= ^, whereinn is the number of reference pixels as well as the number of sample pixels,^^,^^^ is the reference intensity of each reference pixel of the plurality of referencepixels, wherein ^^,^^^is the current intensity of each sample pixel of the plurality of sample pixels, whereinOur ref.: 240259EPWOP 2Sony Semiconductor Solutions Corporation ^^^^=is the mean reference intensity of the plurality of reference pixels, whereinis the mean current intensity of the plurality of sample pixels, wherein ^ =is the standard deviation of the intensities, wherein ^ =is the standard deviation of the current wherein ^^^^ and ^^^^ are predefined correction coefficients.

3. The circuitry of claim 2, wherein the monitoring image is diagnosed as degraded and themonitoring camera as unreliable for monitoring if the noise-adjusted Pearson correlationcoefficient ^^^^ is lower than the predefined threshold.

4. The circuitry of claim 1, wherein the reference intensities are based on the luminancevalues of the plurality of reference pixels and the current intensities are based on the luminance values of the plurality of sample pixels.

5. The circuitry of claim 1, wherein the reference marker region has a reliable contrast,wherein reliability of the contrast is based on a Pearson correlation coefficient of a Pearson correlation between the reference intensities and the current intensities being above a predefined contrast threshold.

6. The circuitry of claim 5, wherein the contrast threshold is at least 0.9.

7. The circuitry of claim 5, wherein the contrast is the greyscale contrast or the colorcontrast.Our ref.: 240259EPWOP 3Sony Semiconductor Solutions Corporation8. The circuitry of claim 1, wherein the circuitry is further configured to determinea mean white intensity of the plurality of sample pixels, and / ora mean black intensity of the plurality of sample pixels, and / ora contrast of the plurality of sample pixels, and / ora blur of the plurality of sample pixels, andwherein diagnosing the monitoring image is further based onthe mean white intensity of the plurality of sample pixels, and / or the mean black intensity of the plurality of sample pixels, and / or the mean black intensity of the plurality of sample pixels, and / or the blur of the plurality of sample pixels.

9. The circuitry of claim 8, wherein the circuitry is further configured to obtaina mean white intensity of the plurality of reference pixels, and / ora mean black intensity of the plurality of reference pixels, and / ora contrast of the plurality of reference pixels, and / ora blur of the plurality of reference pixels,wherein the circuitry is further configured to determine amean white intensity comparison between the mean white intensity of theplurality of sample pixels and the mean white intensity of the plurality of reference pixels, and / or amean black intensity comparison between the mean black intensity of theplurality of sample pixels and the mean white intensity of the plurality of reference pixels, and / or acontrast ratio between the contrast of the plurality of sample pixels and thecontrast of the plurality of reference pixels, and / ora blur ratio between the blur of the plurality of sample pixels and the blur of theplurality of reference pixels, andOur ref.: 240259EPWOP 4Sony Semiconductor Solutions Corporation wherein diagnosing the monitoring image is further based on the mean white intensity ratio and / or the mean black intensity ratio and / or the contrast ratio, and / or the blur ratio.

10. The circuitry of claim 9,^^^^^^^^ wherein the contrast ratio is determined according to^^^^^^^^^^^^^^, wherein the blur ratio is determined according towherein ^^^^^^^^ =^^^^^^^, wherein ^ ^^^^^^^^ =^^^^^^^^^^, whereinis the mean current intensity of the plurality of sample wherein ^ =is the mean reference intensity of the plurality of pixels, wherein ^ =is the standard deviation of the current wherein ^is the standard deviation of the referenceintensities, wherein ^^ ^^^^^^^^^^= ^^^^,^^^, wherein ^^^^,^^^ is the maximum difference in current intensities between adjacent pixelsof the plurality of sample pixels, wherein ^^ ^^^^^^^^^^= ^^^^,^^^, wherein ^^^^,^^^ is the maximum difference in reference intensities between adjacentpixels of the plurality of reference pixels.Our ref.: 240259EPWOP 5Sony Semiconductor Solutions Corporation 11. The circuitry of claim 1, wherein the reference marker region and the marker region are defined based on a virtual marker.

12. A method for diagnosing external interference influencing a monitoring image capturedwith a monitoring camera, comprising: obtaining reference intensities of a plurality of reference pixels of a reference marker region in a reference image captured with the monitoring camera, wherein the reference marker region is smaller than the reference image, obtaining the monitoring image, determining current intensities of a plurality of sample pixels of a marker region in the monitoring image, wherein the marker region corresponds to the reference marker region, and wherein the plurality of sample pixels of the marker region corresponds to the plurality of reference pixels of the reference marker region, determining a correlation between the determined current intensities and the reference intensities, and diagnosing, based on the determined correlation and a predefined threshold, themonitoring image for degradation due to an external interference for evaluating the reliability ofthe monitoring camera as a monitoring tool.

13. The method of claim 12, wherein determining the correlation includes determining anoise-adjusted Pearson correlation coefficient ^^^^ according to the following formula:wherein^ ^^^^^^,^^^= ^, whereinn is the number of reference pixels as well as the number of sample pixels,^^,^^^ is the reference intensity of each reference pixel of the plurality of referencepixels, whereinOur ref.: 240259EPWOP 6Sony Semiconductor Solutions Corporation ^^,^^^ is the current intensity of each sample pixel of the plurality of sample pixels,wherein ^^^^=is the mean reference intensity of the plurality of referencepixels, whereinis the mean current intensity of the plurality of sample wherein ^^^^=is the standard deviation of the reference intensities, wherein ^ =is the standard deviation of the current wherein ^^^^ and ^^^^ are predefined correction coefficients.

14. The method of claim 13, wherein the monitoring image is diagnosed as degraded and themonitoring camera as unreliable for monitoring if the noise-adjusted Pearson correlationcoefficient ^^^^ is lower than the predefined threshold.

15. The method of claim 12, wherein the reference intensities are based on the luminancevalues of the plurality of reference pixels and the current intensities are based on the luminance values of the plurality of sample pixels.

16. The method of claim 12, wherein the reference marker region has a reliable contrast,wherein reliability of the contrast is based on a Pearson correlation coefficient of a Pearson correlation between the reference intensities and the current intensities being above a predefined contrast threshold.

17. The method of claim 16, wherein the contrast threshold is at least 0.9.Our ref.: 240259EPWOP 7Sony Semiconductor Solutions Corporation18. The method of claim 17, wherein the contrast is the greyscale contrast or the colorcontrast.

19. The method of claim 12, wherein the method further comprises determininga mean white intensity of the plurality of sample pixels, and / or a mean black intensity of the plurality of sample pixels, and / or a contrast of the plurality of sample pixels, and / or a blur of the plurality of sample pixels, and wherein diagnosing the monitoring image is further based on the mean white intensity of the plurality of sample pixels, and / or the mean black intensity of the plurality of sample pixels, and / or the mean black intensity of the plurality of sample pixels, and / or the blur of the plurality of sample pixels.

20. The method of claim 19, wherein the circuitry further comprises obtaininga mean white intensity of the plurality of reference pixels, and / or a mean black intensity of the plurality of reference pixels, and / or a contrast of the plurality of reference pixels, and / or a blur of the plurality of reference pixels, wherein the method further comprises determining a mean white intensity comparison between the mean white intensity of the plurality of sample pixels and the mean white intensity of the plurality of reference pixels, and / or a mean black intensity comparison between the mean black intensity of the plurality of sample pixels and the mean white intensity of the plurality of reference pixels, and / or a contrast ratio between the contrast of the plurality of sample pixels and the contrast of the plurality of reference pixels, and / orOur ref.: 240259EPWOP 8Sony Semiconductor Solutions Corporation a blur ratio between the blur of the plurality of sample pixels and the blur of the plurality of reference pixels, and wherein diagnosing the monitoring image is further based on the mean white intensity ratio and / or the mean black intensity ratio and / or the contrast ratio, and / or the blur ratio.

21. The method of claim 20,^^^^^^^^ wherein the contrast ratio is determined according to^^^^^^^^^^^^^^, wherein the blur ratio is determined according to whereinwherein whereinthe mean current intensity of the plurality of sample pixels, wherein ^^^^=is the mean reference intensity of the plurality of reference pixels, wherein ^^^^=is the standard deviation of the current intensities, wherein ^^^^is the standard deviation of the reference intensities, wherein ^^ ^^^^^^^=^^^^^^^,^^^, wherein ^^^^,^^^is the maximum difference in current intensities (24a) between adjacent pixels of the plurality of sample pixels, wherein ^^ ^^^^ =^^^^^^^^^^,^^^, wherein ^^^^,^^^is the maximum difference in reference intensities between adjacent pixels of the plurality of reference pixels.Our ref.: 240259EPWOP 9Sony Semiconductor Solutions Corporation 22. The method of claim 12, wherein the reference marker region and the marker region are defined based on a virtual marker.

Images