SECURE MONITORING OF A PROTECTED AREA USING A TOF CAMERA

Description

[0001] The invention relates to a device and a method for monitoring a protected area. In particular, the invention relates to devices and methods using a ToF (Time-of-Flight) camera.

[0002] In various technical fields, such as industrial production facilities, protective devices are used to protect people from hazards associated with the production process and to protect the production process from disruptions. In addition to protective devices such as barriers with monitored doors, non-contact optical systems such as light barriers are also common.

[0003] Non-contact monitoring can also be achieved by optically detecting a protected area. Conventional digital cameras are known for this purpose, allowing them to capture digital images, i.e., matrix arrangements of individual pixels. Camera systems have also been available for some time that can measure distances using the time-of-flight (ToF) method and assign these distances to image areas. While conventional digital images represent a two-dimensional detection, the distance information from a ToF camera enables a three-dimensional detection of the monitored protected area.

[0004] Known ToF cameras use time-modulated illumination, e.g. with light pulses, whereby the determination of the distance information is based on the time it takes the light to reach the object captured in the respective image area and back to the camera.

[0005] However, known ToF camera systems are complex systems with various optical / electrical components, so that when used in a protective device, the risk of possible malfunctions and resulting incorrect detection must be considered.

[0006] DE 10 2018 116 481 describes a 3D time-of-flight camera for capturing three-dimensional image data from a detection area. An illumination unit is provided for emitting transmitted light modulated at at least two frequencies into the detection area. An image sensor with a plurality of pixels is provided for generating received signals from reflected transmitted light. Evaluation circuits are assigned to the pixels for determining at least one distance per pixel from the received signals using a time-of-flight method based on the two frequencies. An internal reference target directs a portion of the transmitted light or test light within the LED time-of-flight camera onto the image sensor. The 3D time-of-flight camera is designed to be secure and has an evaluation unit that compares a measured variable of the reference target, determined by the evaluation circuits, with a reference measured variable for a functional test.

[0007] US 2019 / 227174 A1 discloses an exposure control device for controlling an integration period of a TOF image capture sensor having an illumination source providing pulsed illumination at a pulse repetition frequency, in which each pixel of an array of pixels is represented by a plurality of pairs of detection values. Each pair of detection values corresponds to detected light sampled according to a pulsed sampling pattern having a respective phase relationship with the pulsed illumination of the illumination source. A detector is configured to determine, for a selected detection value, a set of one or more target pixels, a portion of the detection value that is independent of the integration period and a portion that depends on the integration period.A controller is configured to select a next integration period for the image acquisition sensor such that, for a next acquisition, the portion of the acquisition value that depends on the increased integration period is substantially equal to the difference between the required acquisition value and the portion that is independent of the integration period.

[0008] TheEP 1 825 731 A1 discloses an optoelectronic safety sensor with a light receiver for reliably capturing detection data from a monitored area. A control and evaluation unit is configured to reliably determine its own position from the detection data in order to trigger a position-specific safety-relevant function. The safety sensor is a 3D safety sensor whose detection data is three-dimensional image data. The control and evaluation unit is configured to determine its own position based on three-dimensional features.

[0009] It can be considered a task to propose a device and a method for monitoring the protected area in which safe monitoring can be ensured using a ToF camera.

[0010] The object is achieved by a device according to claim 1 and by a method according to claim 11. Dependent claims relate to advantageous embodiments of the invention.

[0011] The device according to the invention comprises detection means for detecting information from a protected area and evaluation means for evaluating the detected information and for delivering an evaluation signal depending on the detection.

[0012] The various means of the device according to the invention, i.e., the acquisition means and evaluation means, are initially referred to purely functionally here. In suitable embodiments, they may, for example, comprise both hardware and software components. The various means may also be implemented using common components; for example, the same electrical circuits may implement functions of both the evaluation and acquisition means, or software components of both means may be executed on the same computer, processor, or other programmable hardware.

[0013] The protection zone can be any area in which a monitoring function is required, e.g., because a hazard source and / or a facility to be protected is located there. For example, a facility, such as an industrial production plant, may be located within the protection zone, where, in the event of the approach of people and / or objects, a control function can be executed, such as issuing a warning, slowing down, or stopping the facility.

[0014] According to the invention, the detection means comprise at least one ToF camera for capturing digital images of the protected area and for determining distance information from image regions of the digital images of the protected area. A ToF camera is understood to be any device that determines the distances between image regions, i.e., groups of pixels or individual pixels of the digital images, from light-time-of-flight information. The detection means thus provide two-dimensional information as digital images and, in addition, the distance information as a third dimension. The images and distance information are captured in chronological sequence and forwarded to the evaluation means.

[0015] This information is evaluated by the evaluation tools in such a way that people and / or objects within the protection zone are detected. In the simplest case, the detection of objects or people can comprise the mere presence of a body, but optionally also the positioning within the protection zone and / or possible movement and / or identification. In the case of objects, the identification can comprise the detection of the type of object, and in the case of people, an individual assignment and / or membership in a group, e.g., a group authorized to enter certain areas. Image processing algorithms, for example, can be used for detection. The detection takes into account both the two-dimensional digital images and the three-dimensional data from the ToF camera.Detection may be limited to simply detecting the presence of an object or person, but preferably it also includes locating the detected object within the protected area.

[0016] Depending on the detection, the evaluation means can output an evaluation signal, which can be understood as an enable signal, for example, which signals a safe state and can, for example, enable operation of a system within the protected area, and / or an alarm or warning signal, which signals a dangerous state that requires increased attention or a specific response, such as an interruption in operation or activation of a safe state. The positive and negative evaluation signals, i.e. enable signal and alarm / warning signal, can be complementary, i.e. the absence of an enable signal can be understood as a warning signal or vice versa. In the claims and the following description, a positive evaluation signal can therefore always also be understood as the absence of a negative evaluation signal and vice versa.

[0017] According to the invention, the detection means comprise at least one digital image sensor for supplying a sequence of digital images each captured within a controllable integration time. This can be a digital image sensor as part of the ToF camera and / or a separate digital image sensor, e.g. for capturing 2D images in addition to the 3D data of the ToF camera. Any type of image sensor that captures a digital image, i.e., pixels arranged in a matrix, such as, for example, CCD sensors or APS sensors, preferably CMOS sensors, is suitable. In any case, the image sensor has a controllable integration time, i.e., the time within which incident light is captured to capture each image can be adjusted by external control.

[0018] According to the invention, test means are provided for checking the function of the detection means. Like the aforementioned detection and evaluation means, the test means are initially named functionally and, depending on their implementation, can use common hardware and / or software components with the other means of the device. The test means are designed to specify a changed integration time for the image sensor by control and to check an image subsequently acquired with the changed integration time. If the image sensor is functioning properly, a change in the integration time should lead to a changed exposure of the image, i.e., for example, with an extended integration time, higher brightness values should result at individual pixels and / or on average, whereas with a reduced integration time, reduced brightness values should result.

[0019] The test equipment is designed to detect a potential error condition based on the detection of a change in exposure. Depending on the detection of an error condition, the monitoring device can behave differently, in particular, signal the error externally and / or output a negative evaluation signal, even if no people / objects have been detected.

[0020] The determination of whether or not an error condition exists can be signaled by issuing an internal and / or external test signal. This is a purely functional indication, i.e., it does not have to be an individually identifiable electrical signal, but can also be implemented, for example, as a state or variable content in software. The test signal should preferably indicate the proper functioning of the detection means (positive test signal) or, in the event of an error, signal improper functioning (negative test signal). In the claims and the following description, a positive test signal can always also be understood as the absence of a negative test signal, and vice versa.

[0021] The test equipment thus detects whether the exposure of at least one image captured with it within the sequence of images has changed as expected as a result of the changed integration time. If, for example, there is no change, an insufficient change, or even a change in the opposite direction, an error condition can be detected, which can be signaled, for example, by a corresponding negative test signal.

[0022] The device according to the invention thus enables use in safety-related applications. Proper functioning of the detection means with the ToF camera is ensured by at least one, preferably multiple, and particularly preferably cyclical testing, whereby error states can be detected and, if necessary, signaled by a test signal. Thus, for example, the operation of a system within the protected area can be made dependent not only on the signaling of the detection of persons and / or objects, but also on an indication of proper functioning.

[0023] According to a preferred embodiment, the testing means can be designed to determine a deviation in the exposure and to compare the deviations with a deviation threshold. The deviation threshold can thus, for example, indicate a minimum amount of the expected change. If the deviation is above the deviation threshold (or, if applicable, even if the deviation is equal to the deviation threshold), proper function can be determined (test passed) or an error can be detected if a deviation is below the deviation threshold. In the simplest case, a deviation in the exposure can be determined by forming a difference between brightness values. This can, for example, be done at individual pixels or areas of pixels; preferably, values of the majority of the pixels or even all pixels are processed. The exposure of the pixels under consideration can, for example,by calculating a sum and / or average value of the respective brightness. The deviation threshold then specifies the expected minimum value for the change in exposure; if the change is below this value, an error condition must be assumed. Using the deviation threshold ensures a reliable distinction between error and normal cases, so that minor influences such as noise are not mistakenly interpreted as a signal altered by the changed exposure.

[0024] According to the invention, the acquisition means are designed to capture acquisition images with an acquisition integration time. These acquisition images can be processed, for example, by the evaluation means for detecting persons / objects, thus representing, for example, the normal operating case. The test means are designed to specify a modified integration time, referred to as the test integration time. The test integration time is deliberately extended or shortened compared to the acquisition integration time used in other operations.

[0025] The test equipment is designed to compare the exposure of at least one image captured with the test integration time with at least one image captured with the acquisition integration time. If the test integration time is longer than the acquisition integration time, an increased exposure would be expected if the device is functioning properly. In this case, the test equipment is designed to determine proper functioning with a lower exposure of the acquisition image or to detect a fault with a higher or equal exposure of the acquisition image. Conversely, if the test integration time is shortened, a lower exposure is to be expected if the device is functioning properly.In this case, the test equipment is designed to determine proper function at a higher exposure of the captured image or to detect a fault at a lower or equal exposure of the captured image.

[0026] It is preferred to design the test equipment to check the function of the detection means by controlling the image sensor to change the integration time of a plurality of consecutive, temporally discrepant test points. The test is preferably performed continuously at intervals so that proper function is always guaranteed and a fault condition is detected and signaled promptly.

[0027] In addition to checking the function of the recording equipment by changing the integration time, the test equipment can have or implement other test functions. It should first be noted that although other test functions also refer to the same test equipment that carries out a check by changing the integration time, this does not necessarily mean that it always has to be the same hardware / software components. The test equipment can also have different components that carry out the different types of tests. Individual optional test options are listed below, some or all of which can be combined, whereby a determination of proper function is preferably only made if all selected tests are successful, and an error is also detected if only one test is unsuccessful.

[0028] According to a preferred embodiment, testing can be carried out by noise detection. For this purpose, the test equipment can be provided to check the function of the detection means by comparing at least two images taken at different times, e.g. directly one after the other. If differences are detected between the images, proper functioning can be determined, whereas an error is detected if the images completely match. Even without movement of objects in the protected area and with an unchanged arrangement of the detection means, two consecutive images (directly or at a distance) may not be identical if they are functioning properly due to unavoidable noise, e.g. thermal noise.By forming a difference between the values of individual or all pixels, a noise signal must always be detected as at least a very small deviation; otherwise, it would be assumed that the detection equipment has "frozen," which would lead to the detection of an error. The additional verification through noise detection has the advantage that no modification or control of the detection equipment is required, and the captured images can be used without restriction for the detection of persons / objects while they are also being evaluated by the test equipment.

[0029] A further test function can be a comparison of the two-dimensional and three-dimensional data. For this purpose, the test equipment can be provided to test the function of the detection equipment by comparing the digital images with the distance information of image areas. For example, when detecting an object against a background, the two-dimensional image must be recognizable in the pixels, e.g. the contour, while on the other hand the object must stand out from the background in the distance information. Consequently, the test equipment can be designed to determine proper function if there is a match between the digital images, e.g. objects and contours extracted / recognized from them, and the distance information, e.g. contrasts, gradients, etc. extracted from them. In the event of a match orIf the deviation is below a permissible deviation threshold, proper function can be assumed, whereas if the deviation is above a deviation threshold, a discrepancy exists and a fault would be identified.

[0030] A further test function can be the detection of a projected light pattern. For this purpose, controllable projection means are preferably provided to project a light pattern at least into parts of the protected area. The light pattern can be of various shapes and representations, e.g. a grid, but also a single point, a line, a circle, or similar. In this case, the test means are preferably provided to check the function of the detection means by controlling the projection means to project the light pattern, testing at least one digital image to detect the light pattern, and determining proper function or the presence of an error depending on the detection of the light pattern. The projection means can thus be activated, for example, for only a single test or cyclical test. The digital image can be processed to detect the light pattern.If the light pattern is recognized, proper function can be determined, or if the light pattern is not recognized, the fault can be determined.

[0031] Another possible test is a change in the lighting intensity. The detection means can comprise controllable lighting means. The test means are preferably provided for checking the function of the detection means by controlling the lighting means to change the lighting intensity and to test a changed exposure of at least one image. For example, an increased or, alternatively, a reduced lighting intensity can be specified and an image taken with the changed exposure can be compared with a previous, unchanged exposure. If the lighting intensity is increased for testing purposes, an increased exposure should be recognizable under normal operating conditions, or if the lighting intensity is reduced, a lower exposure should be recognizable.Depending on the detection of a change in exposure as a result of the change in the intensity of the illumination, a corresponding determination can be made.

[0032] Another way of testing the proper functioning of parts of the device can be by observing specific areas of the captured images. For this purpose, areas within the captured images that are not variable during operation should preferably be observed, e.g. a ceiling / wall area in which no people or objects are expected to be present. The testing equipment can have a memory for storing a reference image and defining reference areas therein, whereby the above-mentioned non-variable areas can preferably be used as reference areas. The definition of reference areas can be automatic or manual. The testing equipment can be provided to check the function of the detection means by comparing reference areas of at least one image with the reference areas of the reference image.When considering the preferred non-variable areas, proper function would require at least a high degree of agreement, meaning that, apart from minor deviations such as noise, an identical representation of the reference areas in the reference image and the currently acquired images would be evident. The test equipment is designed to determine agreement or differences and, depending on this, can identify a fault or proper function. In the case of significant differences, a fault would be assumed, and proper function would be assumed if the agreement or deviations are only below a minimal threshold (e.g., noise).

[0033] The device can react differently to the detection of people and / or objects in the safety zone. For example, the detection of people / objects may always result in the output of a negative evaluation signal if no objects or people are accepted in the protection zone. However, it may also be preferable to treat different zones differently. For example, one or more safety zones and / or one or more warning zones can be defined within the captured images or within the protection zone. These zones can be defined automatically or manually.Preferably, the evaluation means can be configured to output a negative evaluation signal, i.e., a security alarm signal, upon detection of persons and / or objects in the security zone, and / or to output the evaluation signal as a warning signal upon detection of persons and / or objects in the warning zone. This makes it possible to respond flexibly to the detection of persons or objects.

[0034] According to preferred embodiments, the ToF camera comprises a controlled illumination device for emitting the modulated light, an image capture device for capturing digital images of the protected area, and a capture time evaluation device for determining distance information of image areas of the digital images of the protected area based on delay times between the emission of the modulated light and the capture of the image areas.

[0035] The invention further relates to a method for monitoring a protected area, in which digital images and distance information from image areas of the digital images of a protected area are captured using a ToF camera, the digital images and distance information are evaluated to detect persons and / or objects within the protected area, and an evaluation signal is output depending on the detection. To enable a safety-related application, a functional test is performed by controlling an image sensor used to capture the digital images to change an integration time, checking a changed exposure with at least one image, and outputting a test signal depending on the detection of a changed exposure.

[0036] The following are examples of embodiments described in more detail with the aid of drawings. These show: Fig.1 shows a schematic, perspective view of a first embodiment for monitoring a protected area by means of an optical protective device; Fig.2 shows the optical protective device from Fig. 1 and objects arranged in front of it in a schematic plan view; Fig. 3 a block diagram of the optical protection device from Fig. 1, Fig. 2 ; Fig. 4 shows a schematic, perspective view of a second embodiment for monitoring a protected area by means of an optical protective device.

[0037] In various areas, such as industrial production facilities, protective devices can protect people from hazards associated with production processes or prevent disruptions to the production process. The following examples of area monitoring systems are presented, which prove advantageous over guards, particularly in areas where human-machine interaction is necessary.

[0038] Fig. 1 As a first application example, it shows the monitoring of a protected area 10 in which a first machine 14, for example, an industrial robot, is located. A second, mobile machine 16, for example, an autonomous industrial truck, moves through the protected area 10.

[0039] An optical monitoring device 20 is mounted stationary so that it optically detects the protected area 10 comprising the first, stationary machine 14 and the second, mobile machine 16. Within the protected area 10, a safety zone 12 is defined around the first machine 14.

[0040] A control device 22 controls the first machine 14 and, via a wireless connection (not shown), also the second machine 16. The control device 22 is coupled to the optical monitoring device 20 and receives an evaluation signal A from it, for example, via an OSSD output. As explained in more detail below, the evaluation signal A indicates whether or not persons or objects are currently being detected in the security area 12. At the same time, the evaluation signal A also indicates the proper functioning of the optical monitoring device 20.

[0041] As long as the evaluation signal A indicates in the form of a release signal that the safety area 12 is clear and, in addition, the proper function of the optical monitoring device 20 is ensured, the control device 22 controls the operation of the machines 14, 16 according to the respective work sequence. Persons 24 are protected from the first machine 14 as long as they are located or moving outside the safety area 12.

[0042] If the optical monitoring device 20 detects the presence of a person or an object in the safety area 12 as explained below and communicates this to the control device 22 by means of a corresponding evaluation signal A in the form of an alarm or warning signal, the control device 22 stops further operation of the machines 14, 16 or puts them into a safe state so that people are protected and collisions with objects are avoided.

[0043] Within the safety area 12, a special feature applies to the second machine 16. A cutout area 18, ie a section of the safety area 12, is defined around it, which allows the second machine 16 to move within the safety area 12 in coordination with the control device 22 without this leading to a warning signal.

[0044] The optical monitoring device 20 comprises a Time-of-Flight (ToF) camera, which records both a conventional digital 2D color image and a 3D image. Fig. 2shows a schematic of the arrangement of a person 24 and objects 26a, 26b in the protection zone 10 detected by the optical monitoring device 20. The 3D image is created using a special light source and exposes the image in the non-optical range in such a way that a special sensor array can measure the time of flight of the photons in order to determine the respective distances d1, d2, d3 of the person 24 and the objects 26a, 26b from d = ct / 2. Corresponding light source-detector combinations are available in various designs, for example, to measure the time of flight using phase shift or extremely fast electronics.

[0045] Fig. 31 shows a block diagram of the monitoring device 20 as an example of an embodiment. This comprises, as the detection means, a ToF camera module 30 with a 2D image sensor 32 and a 3D sensor 34 according to the functional principle described above. The ToF camera module 30 further comprises a microcontroller 36, a power supply 38, and a communication module 40 with high-speed capacity.

[0046] Also part of the monitoring device 20 is an illumination 42 for the 2D sensor 32, an illumination 44 for the 3D sensor 34, and a laser module 46 for projecting a light pattern as a reference, as explained later. The illumination 42 can, for example, be in the visible or near-infrared range to avoid irritation. The illumination 44 for the 3D sensor is modulated, usually in the near-infrared range. Depending on the implementation of the ToF camera, a phase shifter or ultrashort pulses are used.

[0047] The monitoring device 20 comprises an evaluation unit 50 with microelectronics, preferably with at least two processing cores. A pre-evaluation module 52 and a safety module 54 are implemented in hardware and / or software as evaluation and testing means, as explained below. Furthermore, the evaluation unit 50 has a power supply designed for safety-related devices (not shown).

[0048] The pre-evaluation module 52 is connected to the ToF camera module 30 via a first I / O electronic module 56. The safety module 54 is connected to the laser module 46, the ToF camera module 30, and an external bus connection 60, e.g., a safety-related CAN bus implementation, via a second I / O electronic module 58.

[0049] The safety module 54 is connected to the pre-evaluation module 52 and has a safety-related redundant and monitored processor pair 62 and a communication chip 64. This means that one part is monitored by another part in order to exclude errors.

[0050] During operation, the ToF camera module 30 cyclically delivers two images: a 2D image captured by the 2D image sensor 32 and a 3D image captured by the 3D sensor 34 under control of the illumination 44. The 2D image sensor captures one image each in the optical and NIR ranges.

[0051] The 2D and 3D image data are processed in the evaluation unit 50, initially by the pre-evaluation module 52. This contains microelectronics that performs pre-processing. This prevents the latency from being unnecessarily increased. The pre-evaluation module 52 and / or the downstream safety module 54, as evaluation tools, evaluate the 2D and 3D image data using digital image processing methods by detecting objects and / or people and locating them within the protection zone 10. The identification and, in particular, the localization of detected objects / people is based precisely on the 3D data, which, on the one hand, allows a three-dimensional shape / contour of objects / people to be recorded and, on the other hand, allows their position to be determined with high reliability.The evaluation means in the pre-evaluation module 52 and / or the safety module 54 check whether there are any objects / persons within the safety area 12 (whereby the mobile machine 16 in the cutout area 18 is ignored). As long as the evaluation shows that there are no persons / objects within the safety area 12, the evaluation signal A is output, for example, via an OSSD output (not shown), as an enable signal; otherwise, it is output as a warning signal.

[0052] The safety module 54 also implements test equipment to ensure the proper functioning of the components of the monitoring device 20. To this end, the safety module cyclically performs a number of different tests to verify the functionality of the components. The test results can be summarized in a test signal P.

[0053] The tests performed by the safety module include the following procedures. Depending on the application, individual, some, or all of the described test procedures can be used. According to the invention, at least the following test is used by changing the integration time. Test by projecting a light pattern

[0054] During this test, the safety module 54 controls the laser module 46 to project a light pattern as a reference image into the detected protection zone 10. The light pattern can, for example, comprise one or more individual points, but can also, for example, comprise a grid of lines. The safety module 54 checks whether the light pattern projected as a reference is contained in the 2D image supplied by the ToF camera module 30. If no sufficient match is found, an error is detected, i.e., a negative test signal P is generated. Test by noise analysis

[0055] For this test, the safety module 54 compares consecutive 2D images delivered by the ToF camera module 30. A point-by-point difference is calculated between the images. Even in a stationary scenario within the protection zone 10, the images will differ due to noise, so the difference must be different from zero for a proper signal. In the case of a complete match, a "frozen" memory is assumed and an error is detected, e.g., a negative test signal P is generated. Test by changing the integration time

[0056] The safety module 54 controls the ToF camera module 30 so that the integration time of the 2D image sensor 32 is changed. The exposure of an image acquired with the changed integration time is then compared with that of an image acquired before or after with the unchanged integration time. A difference in the brightness values is calculated and compared with a minimum threshold. Extending the integration time must result in an exposure increased by at least the specified threshold; reducing the integration time must result in a correspondingly reduced exposure. If this condition is not met, an error is detected, e.g., a negative test signal P is generated. Test by changing the lighting intensity

[0057] For this test, the safety module 54 controls an illumination source, e.g., the illumination source 42 for the 2D sensor, such that the illumination intensity is changed. For verification, the exposure of an image captured with the changed illumination intensity by the 2D image sensor 32, the 3D image sensor 34, or both, is compared with that of an image captured before or after with the unchanged illumination. As with the test by changing the integration time, a difference in the brightness values is calculated and compared with a minimum threshold. If the expected increased or decreased exposure does not result, a malfunction in the illumination source and / or the image sensor is detected, and, for example, a negative test signal P is generated. Test by comparison with reference image

[0058] After the system is installed, a reference image of the image area is captured. This reference image stores a set of measured values under operating conditions in the acquisition unit, covering the illumination and geometry of the image area. This represents the safe situation in which safety is guaranteed.

[0059] During operation, the measured distances are compared with the reference recording at regular intervals in areas where this is possible to ensure that the distance measurement is functioning properly. Deviations above a specified threshold are identified as an error. Test by comparing 2D and 3D data

[0060] In this test, the 2D image is compared with the 3D image. If there are deviations in the representation above a specified threshold, which, for example, represents the measurement uncertainty, a fault has occurred in one of the sensor modules, and a fault is identified.

[0061] The entire system is programmed using external software, which programs the evaluation unit, for example, via USB. The software can define safety zones and warning zones, for example, and define the functions of the inputs and outputs within a corresponding framework. External signals, for example, can change the position and presence of protected zones (e.g., with regard to the cutout area).

[0062] If one of the tests detects a safety risk, which is indicated here, for example, by a negative test signal P, this is signaled externally via the interfaces, e.g., CAN bus 60. In addition, a negative evaluation signal A is output, since the safe state cannot be guaranteed.

[0063] It should be noted that references to the evaluation signal A and the test signal P are to be understood here in a logical / functional sense, and in an implementation these do not necessarily have to be present as separate electrical signals, but also, for example, as logical states or stored information, e.g. flags in a software implementation.

[0064] Fig. 4As a second embodiment, it shows a protected area 110 with two mobile machines 114, 116, which are equipped with optical monitoring devices 20 and each monitor a safety area 112. A cutout area 118 is excluded from this.

[0065] The protected area can, for example, have a different geometry at different heights, which may also be adaptive. 3D recognition makes it possible to identify people. Certain areas in industrial environments are often restricted to trained personnel. If this is detected, production can continue as planned despite the presence of a person. If an unauthorized person enters the area, appropriate precautions can be taken.

[0066] In summary, the exemplary embodiments explained enable devices and methods for monitoring a protected area with a ToF camera, in which test equipment is provided to verify the function of the detection equipment. The detection equipment, as well as the test equipment, can be designed differently in various embodiments, even deviating from the above-mentioned examples. For example, while two separate illumination systems 42, 44 are provided above, additional illumination systems can also be provided, or instead of two, only one common illumination system. Likewise, instead of separate 2D and 3D image sensors 32, 34, only a single image sensor can be provided, whose signal is processed.

[0067] In any case, the test equipment detects a possible error condition, so that the evaluation signal to be output is safeguarded.

Claims

1. Device for monitoring a protective area, comprising - acquisition means (30), at least comprising a ToF camera for acquiring digital images of the protective area (10) and for determining distance information of image areas of the digital images of the protective area (10), - evaluation means (50) for evaluating the digital images and the distance information for detecting persons and / or objects within the protective area (10), - the acquisition means (30) have at least one digital image sensor (32) for supplying a sequence of images acquired within a controllable integration time in each case, the acquisition means (30) being designed to acquire acquisition images with an acquisition integration time, - and verification means (54) are provided for verifying the function of the acquisition means (30) characterized in that - the verification means (54) are designed for specifying a modified verification integration time, namely one which is longer or shorter than the acquisition integration time, - and for comparing the exposure of at least one image acquired with the verification integration time with at least one acquisition image, - wherein, in the case of an extended verification integration time, the verification means (54) are designed to detect a fault state with a larger or equal exposure of the acquisition image, - and wherein, in the case of a shortened verification integration time, the verification means (54) are designed to detect a fault state with a smaller or equal exposure of the acquisition image.

2. Device according to claim 1, wherein - the verification means (54) are designed to determine a deviation of the exposure and to compare the deviation with a deviation threshold.

3. Device according to one of the preceding claims, wherein - the verification means (54) are designed to carry out the verification of the function of the acquisition means (30) by controlling the image sensor to change the integration time at a plurality of verification times spaced apart in time.

4. Device according to any one of the preceding claims, wherein - the verification means (54) are further provided for verifying the function of the acquisition means (30) by comparing at least two images and detecting a fault state when the images completely match.

5. Device according to any one of the preceding claims, wherein - the verification means (54) are further provided for verifying the function of the acquisition means (30) by comparing the digital images with the distance information of image areas for detecting a fault state in case of a deviation above a deviation threshold.

6. Device according to one of the preceding claims, wherein - controllable projection means (46) are provided for projecting a light pattern at least into parts of the protective area (10), - and the verification means (54) are further provided for verifying the function of the acquisition means (30) by controlling the projection means (46) for projecting the light pattern, verifying at least one digital image for detecting the light pattern and detecting a fault state depending on the detection of the light pattern.

7. Device according to one of the preceding claims, wherein - controllable illumination means (42, 44) are provided - and the verification means (54) are further provided for verifying the function of the acquisition means (30) by controlling the illumination means (42, 44) to change the intensity of the illumination and for verifying a changed exposure of at least one image and detecting a fault state depending on the detection of a changed exposure due to the change in the intensity of the illumination.

8. Device according to any one of the preceding claims, wherein - the verification means (54) comprise a memory for storing a reference image and a determination of reference areas therein, - and the verification means (54) are further provided for verifying the function of the acquisition means (30) by comparing reference areas of at least one image with the reference areas of the stored reference image and detecting a fault state depending on the detection of differences.

9. Device according to any one of the preceding claims, wherein - at least one safety area (12) is defined within the protective area (10), wherein the evaluation means (50) are designed such that an evaluation signal A is output as a safety alarm signal when persons and / or objects are detected in the safety area (12) and / or - at least one warning area is defined within the protective area (10), the evaluation means (50) being designed in such a way that an evaluation signal (A) is output as a warning signal when persons and / or objects are detected in the warning area.

10. Device according to one of the preceding claims, wherein the TOF camera comprises - a controlled illumination device (44) for emitting modulated light, - an image acquisition device (34) for acquiring digital images of the protective area, and - an acquisition time evaluation device for determining distance information of image areas of the digital images of the protective area on the basis of delay times between the emission of the modulated light and the acquisition of the image areas.

11. Method for monitoring a protective area, in which - digital images of a protective area (10) and distance information of image areas of the digital images of the protective area (10) are acquired by means of a TOF camera (30) by recording acquisition images with an acquisition integration time, - the digital images and distance information of the acquisition images are evaluated for detecting persons and / or objects within the protective area (10) characterized in that - a functional verification is carried out by specifying a modified verification integration time, namely a verification integration time which is longer or shorter than the acquisition integration time, and recording images with the verification integration time - and by comparing the exposure of at least one image acquired with the verification integration time with at least one acquisition image, - wherein, in the case of an extended verification integration time, a fault state is acquired with a larger or equal exposure of the acquisition image, - and wherein, in the case of a shortened verification integration time, a fault state is detected with a smaller or equal exposure of the acquisition image.