Cavity inspection device and method for controlling the illumination of a cavity inspection device

Abstract

Cavity inspection device with at least one camera (10, 12) movable in a cavity (4), a control device (13) connected to the camera (10, 12) and an illumination device (14, 16) controllable by the control device (13), characterized in that the control device (13) has an image analysis device which is designed such that it recognizes at least one surface property in an image recorded by the camera (10, 12) which has an influence on the reflection and illumination, and controls exposure settings of the camera and / or the illumination device (14, 16) on the basis of this recognized surface property (18).

Description

The invention relates to a cavity inspection device and to a method for controlling the illumination of a cavity inspection device.Cavity inspection devices are used, for example, to inspect pipelines, containers or shafts from the inside with a camera. In particular, a cavity inspection device can be a sewer pipe inspection device with which sewer pipes of sewers are optically inspected. Such cavity inspection devices generally have an illumination device and a camera. It is known to control the illumination in order to achieve a desired exposure and to prevent reflections which interfere with the recorded image. In such a control, fluctuations or changes in the illumination properties occur, which impair the image recording. DE 10 2018 212 216 A1 discloses a cavity inspection system having an image evaluation device which is designed for the automated recognition of structures in the image. Based on this image recognition, a camera can be pivoted in order to bring recognized structures larger into the image. In this case, automated illumination adaptation is disclosed for closer observation of the structures.Against this background, it is an object of the invention to improve a cavity inspection device or the operation of a cavity inspection device to the effect that illumination which is better adapted to the environmental conditions is realized, which illumination reacts in particular better to changes in the environmental conditions.This object is achieved by a cavity inspection device having the features specified in claim 1 and by a method for controlling the illumination of a cavity inspection device having the features specified in claim 11. Preferred embodiments are evident from the dependent claims, the following description and the attached figures.The cavity inspection device according to the invention serves for inspecting cavities as they are present in containers, tanks, pipelines, shafts or sewer pipes. The cavity inspection device is particularly preferably an inspection device for sewer pipes, i.e. sewers. The cavity inspection device has at least one camera movable into a cavity. The camera can be designed, for example, as a sliding camera, which is moved in the cavity, for example a canal tube, with the aid of a sliding rod. Alternatively, it can be an inspection robot, for example, which has wheels and travels in a sewer pipe, for example. In a further possible embodiment, the at least one movable camera can also be arranged on a flying inspection device, such as a drone. The inspection device could also be designed as an inspection device which can be moved vertically with the aid of a cable winch. The cavity inspection apparatus has a control device connected to the camera and an illumination device controllable by the control device. The control device and illumination device are preferably designed such that the control device controls the illumination in order to achieve a desired illumination of the camera image. According to the invention, the control device has an image analysis device which is designed in such a way that it identifies at least one surface property in an image recorded by the camera and controls the exposure settings of the at least one camera and / or the illumination device on the basis of the identified surface property. This means that, according to the invention, the control of the illumination and / or adjustment of the exposure of the at least one camera is carried out not or not only by measuring the image brightness or measured reflected light intensities, but by recognizing structures or surfaces in the recorded image. For example, the image analysis device can be designed such that it can distinguish smooth from rough surfaces, detect moist surfaces or water surfaces and the like. The surface properties are surface properties which have an influence on the reflection or illumination. This may also include the brightness of the surfaces and the like. The embodiment according to the invention has the advantage that predictive exposure and / or illumination control becomes possible. If certain surface properties are recognized in the image, the illumination and / or the exposure can be adjusted before incorrect illumination, for example glare or too low illumination, occurs. In this respect, undesired brightness fluctuations or cross-fades in the image are prevented.The image analysis device preferably comprises an artificial intelligence system. This can be a computer system with a corresponding software application. The artificial intelligence can have a neural network, for example, and can be trained, for example, by deep learning in order to recognize different surface properties in an image recorded by the camera. Such an artificial intelligence system can be further trained and optimized, according to a further preferred embodiment can also be further trained and optimized during the running operation of the cavity inspection device.The control device with the image analysis device can be integrated into the camera or into an inspection device having the camera. Alternatively, parts of the image analysis device or the entire image analysis device can be arranged at a distance from the camera or from an inspection device carrying the camera and can be connected to the inspection device or the camera and the illumination device via a data connection. The control device can be arranged, for example, outside a channel tube and be connected via a cable to an inspection device which carries the camera. It is also possible to operate the image analysis device or parts thereof in a cloud computer system.According to a further possible embodiment, the cavity inspection device has at least one movement device connected to the camera, which is configured in such a way that it can move the camera. This can be, for example, a pivot drive which pivots the camera. Furthermore, a movement device can also be provided which moves the camera in the cavity, for example moves it back and forth in a canal tube. Such a movement device can be designed, for example, in the form of a carriage or a sliding rod.The control device is preferably designed such that it takes into account movement information about a movement, in particular a future or intended movement of the camera, when controlling the illumination device and / or the exposure settings of the at least one camera. The exposure and / or the illumination can thus be controlled or adapted in a predictive manner. Preferably, the at least one movement device is controllable by the control device or the control device communicates with a movement controller which controls the movement device. By means of this combination, the control device is provided with information about the direction in which the at least one camera is moved by the movement device. This information can be used together with the surface properties detected by the image analysis device for exposure and / or illumination control in such a way that a predictive exposure and / or illumination adaptation takes place.According to a further possible embodiment, the control device can be configured such that it controls the exposure settings and / or the illumination device on the basis of a recognized surface property which, according to taken into account movement information in a subsequent camera position, will be essential for the image exposure. If the control device knows how the viewing direction of the camera will change or change, it can predictively adapt the exposure and / or illumination to those surface properties which will dominate the field of view of the camera in a next step. Thus, for example, a surface region which can initially only be seen in the edge region of an image can get into the center of the image as a result of the movement of the camera and the illumination in the image can then dominant there. If, for example, a wet region initially recognizable only at the edge of the image reaches the center of the image, the reflection will become stronger and the exposure and / or the illumination can be adapted accordingly in a predictive manner for compensation.The control device is further preferably configured such that it takes into account the configuration of the illumination device and the properties of the camera, for example in such a way that it has learned, or by means of artificial intelligence, in which image regions reflections can lead particularly to disturbances. In addition, the image analysis device can be designed such that it also recognizes and takes into account angular positions or alignments of surfaces and that the viewing direction of the camera is taken into account by the control device during the exposure and / or illumination control. Thus, the control device can calculate from the angles of the surfaces and viewing direction of the camera and known radiation angles of the illumination device whether and where reflections which could interfere with the image are to be expected. The control device can then predictively adapt the setting of the exposure and / or illumination by correspondingly controlling the camera and / or the illumination device.Further preferably, the control device and the illumination device are configured such that the brightness of the illumination device and / or the light temperature and / or the emission characteristic of the illumination device can be changed. The brightness can be changed, for example, by dimming individual light sources and / or by switching individual light sources on and off. The light temperature can be changed by changing the light temperature of individual light sources or likewise by gating and blocking out individual light sources. The emission characteristic could be changed by moving individual light sources, adjustable optics or else by switching light sources of different directions on and off. By changing such illumination settings, it is possible to react to the reflection properties of surfaces which are to be viewed or the illumination can be adapted to the object to be observed in such a way that, for example, disturbing reflections are avoided as far as possible.According to a further possible embodiment of the invention, the control device and the camera can be configured such that the control device can change the aperture, shutter time and / or filter settings of the camera. By varying the aperture and shutter speed, the exposure can be adjusted in the usual manner. Alternatively or additionally, filters can be placed in the beam path of the camera and / or their settings can be changed in order to adapt the exposure. This can be carried out alternatively or in combination with the illumination settings described above in order to achieve optimum exposure and sharpness of the images recorded by the at least one camera.According to a further possible embodiment of the invention, the control device is designed such that the surface property defines a reflection property. Alternatively, the control device can preferably be designed such that it assigns reflection properties or exposure and / or illumination settings to detected surface properties, for example on the basis of reflection properties and / or exposure / illumination settings stored in a table or database. Thus, exposure and / or illumination settings can be assigned to specific surface properties in a database or reflection properties, for which corresponding exposure and / or illumination settings are in turn stored in a database. This enables the control device to adjust the illumination and / or exposure in a targeted manner on the basis of recognized surface properties or surface structures.The exposure and / or illumination settings assigned to the reflection properties and / or the surfaces can be permanently predefined in the system. According to a special embodiment of the invention, the control device can be designed such that it automatically determines and adapts depending on the surface properties and, in particular, reflection properties, exposure and / or illumination settings associated with unknown surface properties. For this purpose, the control device can use an artificial intelligence, which realizes a self-learning exposure and / or illumination setting as a function of recognized surface properties. The reflection properties, exposure and / or illumination settings for specific surface structures or surface properties can be determined by the control device by measurement. Brightness settings can thus be varied, for example, and the control device can determine, via the connected camera, at which exposure and illumination setting the best image quality is achieved. This exposure and / or illumination setting can then be assigned to a surface property and preferably stored, for example stored in a data memory of the control device in a table or database, in order later to rely on these settings when such surface properties are recognized.According to a further possible embodiment, the control device can be connected to two or more cameras, wherein the cameras each have an associated illumination device which can be controlled by the control device. The control device is preferably designed such that the image analysis device recognizes surface properties in at least one image of at least one first camera and that the control device controls the exposure setting and / or illumination device of at least one second camera on the basis of these surface properties recognized in the image of the first camera. The at least two cameras preferably have different, preferably non-overlapping, viewing directions, so that when the cameras are moved together, the first camera preferably sees a specific surface structure or surface property before the second camera sees the same location. The two cameras are preferably arranged on a common inspection device and are moved jointly through the cavity, for example a duct tube. In this embodiment, the first camera can thus first recognize the structures to which the exposure and / or illumination must be adapted, so that the exposure and / or illumination for the second camera is preferably always correctly set in advance. Such an arrangement is particularly advantageous in a spherical image scanner which has two cameras with fisheye lenses and viewing directions facing away from one another. An inspection device can have, for example, a first camera at the front end in the feed direction and a second camera at the rear end in the feed direction. During the feed, the first camera thus always sees objects or structures first, so that the exposure setting and / or the illumination control of the illumination device of the rear second camera can be controlled or set on the basis of the image of the first camera. The time of the change in exposure and / or change in illumination or adjustment of illumination is preferably set by the control device as a function of the distance between the cameras and the speed of movement of the cameras.In addition to the above-described cavity inspection device, the subject matter of the invention is a method for controlling the exposure and / or illumination of a cavity inspection device, in particular a cavity inspection device, as has been described above. Preferred features of the cavity inspection device as described above are likewise to be regarded as preferred features of the method described below, and vice versa.According to the method for controlling the illumination and / or exposure setting of a cavity inspection device, a camera image is captured. In a subsequent step, the camera image is analyzed with regard to the surface properties or surface structures in individual image regions. This is preferably done with the aid of an artificial intelligence. The different surface properties can result, for example, from the structure of the surface, e.g. of the inner wall of a duct tube, but can also be influenced by moisture, dirt, encrustations or the like. Bright inliners in a canal pipe can also lead to strong light reflections. Furthermore, structures such as protrusions, depressions, branching lines or the like can influence the surface properties. Surface properties are preferably considered which have an influence on the illumination, that is to say in particular have different reflection characteristics, different depths and / or different brightnesses. In a subsequent step, according to the method, the exposure and / or illumination are adapted to the surface properties determined in the camera image, for example brightness, emission directions and the like of at least one illumination device are adjusted or controlled in a suitable manner. Alternatively or additionally, for example, the diaphragm and exposure time of the camera can be adjusted in a suitable manner, wherein for example desired sharpness adjustments can also be taken into account.Preferably, in a subsequent step, it can be analyzed whether in a subsequent step, a movement of the image section is to take place, for example by pivoting the camera or moving the inspection device, which leads to regions with changed surface properties entering the image section in an amplified manner. Subsequently, the exposure and / or the illumination are adapted to the changing surface properties before or during the movement of the image section towards the region with changed surface properties. This means that the exposure and / or the illumination are adjusted ahead, so that it is ideally correctly adjusted when the image section views a specific area of the object to be examined.Further preferably, in a next step after the movement has taken place, the exposure and / or the illumination in the image recorded by the camera can be checked and, if appropriate, automatically or manually adjusted once again if these are incorrect, if, for example, there is an overexposure or underexposure in the image or individual image regions. For this purpose, too, the control device can be configured in a suitable manner. Preferably, the control device and more preferably its artificial intelligence is trained simultaneously in such a way that, if appropriate, the exposure and / or illumination settings which are assigned to the recognized surface properties are corrected on the basis of the actual setting or adaptation of the exposure and / or illumination setting. This can also be done in the case of surface structures or surface designs which are not yet known to the control device. For such, the exposure and / or the illumination can be varied until the optimum setting is found, which can then be assigned to the detected surface structure.The invention is described below by way of example with reference to the attached figures. In these, the following shows: FIG. 1 schematically shows an example of a sewer pipe inspection system according to the invention, FIG. 2 schematically shows a first camera image, and FIG. 3 schematically shows a second camera image.FIG. 1 shows an example of a cavity inspection system of a sewer pipe inspection system. This has an inspection device in the form of a carriage 2 which is designed to pass through a duct 4. In this example, the car 2 is connected to an external control device 8 via a cable 6. The vehicle 2 could, however, also be designed to be autonomously driving, wherein the control device or its functionality can be at least partially integrated into the vehicle. In such a case, the cable 6 could be dispensed with and the carriage 2 could be equipped with a battery for supplying energy. Instead of a carriage 2, for example, an aircraft-capable inspection device in the manner of a drone could also be used. In the example shown, the carriage 2 has a first camera 10 at a front end and a second camera 12 at its rear end. The first camera 10 is looking forward in the feed direction S, while the second camera 12 is looking rearward in the opposite feed direction S. The cameras 10, 12 in the car 2 are connected to a control device 13. A first lighting device 14 directed forward is arranged on the first camera 10. A second illumination device 16 is arranged on the second camera 12, which second illumination device radiates rearward in the direction of the viewing direction of the second camera 12. The lighting devices 14 and 16 can be formed by an arrangement of a plurality of lighting means, such as light-emitting diodes, surrounding the cameras 10 and 12. The first illumination device 14 and the second illumination device 16 are likewise connected to the control device 13 for controlling the illumination.The control device 13 contains a computer system with an artificial intelligence which is designed for the analysis of the images recorded by the cameras 10 and 12. Thus, the control device 13 can in particular analyze the image recorded by the front camera 10 as to which surface structures are located in the field of view of the camera 10. Depending on these surface structures, the control device 13 controls the illumination devices 14, 16, for example in their brightness or emission direction. To change the emission direction, different lighting means can be switched on or off or the emission characteristic can be changed via adjustable optics. Furthermore, the control device 13 can be configured such that it also controls the exposure settings of the cameras 10 and 12, e.g. suitably sets the aperture and / or the exposure time. When the carriage 2 moves in the advancing direction S, the first camera 10 first sees objects and surfaces in the duct 4. By evaluating the image of the camera 10, it can be recognized whether surfaces in the image change during the movement such that an adaptation of the exposure and / or the illumination is necessary, for example because reflection properties change. This makes it possible to adjust the second illumination device 16 and / or the exposure settings of the second camera 12 in advance such that the exposure and the illumination are set to be equally correct when a corresponding object or a corresponding surface region reaches the field of view of the second camera 12. Such a surface modification or structure could be, for example, a puddle 18 on which light reflections can occur. The closer the first camera 10 comes to this puddle 18, the stronger its influence on the image becomes. Thus, the illumination by the first illumination device 14 and the exposure settings of the first camera 10 can likewise be adapted such that this influence is compensated for or the exposure and illumination are suitably adjusted.This will be described in more detail with reference to FIGS. 2 and 3, which schematically show images as might be recorded by the first camera 10. The first image according to FIG. 2 shows the puddle 18 at a great distance, that is to say from a position of the carriage 2 which is further spaced apart from the puddle 18. In this position, the light reflection which occurs at this puddle 18 still has little influence on the overall illumination of the image. However, the puddle 18 can already be identified as a modified surface structure by the control device 13 during the image analysis in the image. The control device 13 of the carriage 2 also has information from the drive of the carriage 2 about the feed speed or is connected to the external control device 13 from which it receives this information. Thus, the controller 13 may predict when the puddle 18 is so far in the image that it more dominated the image area and has a more influence on illumination or light reflections. Such a state is shown in FIG. 3. By way of the prediction, the control device 13 can adjust the first illumination device 14 and / or the exposure settings of the first camera 10 in a leading manner and adjust the exposure and the illumination such that it is adjusted to the increasing proportion of the puddle or surface structure 18 on the image region.The optical properties or reflection properties of specific surface structures under surface conditions of the channel tube 4 can be stored in a data memory 20 of the control device 13. The exposure and illumination setting can furthermore be continuously readjusted via the current image of the camera 10 and / or of the camera 12. These readjustments can be used for training the artificial intelligence of the control device 13, that is to say the control device 13 can adapt or correct corresponding exposure and / or illumination settings which are stored in the data memory 20 for specific surface conditions. Furthermore, for new surface structures not previously recognized, a corresponding exposure and / or illumination setting can be stored in an automated manner.Instead of integrating the control device 13 and the data memory 20 into the carriage 2 or an inspection device, the control device 13 and / or the data memory 20 or its functionality could also be arranged wholly or partly in an external computer system such as the external control device 8, or the functionality could be provided wholly or partly by a cloud computing system. For this purpose, the external control device and / or the inspection device can be connected to the Internet.According to the invention, therefore, by image analysis of the images recorded by the cameras, a leading exposure and illumination adjustment is made possible, which makes it possible to adjust the exposure and illumination to changes in the image at the moment these changes actually occur. For this purpose, the movement of the cameras 10 and 12 is taken into account by the control device 13.List of reference characters2 Trolley 4 Sewer pipe 6 Cable 8 External control device 10 First camera 12 Second camera 13 Control device, artificial intelligence 14 First lighting device 16 Second lighting device 18 Puddle, surface properties 20 Data memory S Advancing device

Claims

Cavity inspection apparatus having at least one camera (10, 12) movable in a cavity (4), a control device (13) connected to the camera (10, 12) and an illumination device (14, 16) controllable by the control device (13), characterized in that the control device (13) has an image analysis device which is designed such that it identifies at least one surface property which has an influence on the reflection and illumination in an image recorded by the camera (10, 12), and controls exposure settings of the camera and / or the illumination device (14, 16) on the basis of this identified surface property (18).Cavity inspection device according to claim 1, characterized in that the image analysis device comprises an artificial intelligence system.Cavity inspection device according to claim 1 or 2, characterised byat least one movement device connected to the camera (10, 12) for moving the camera (10, 12), which movement device is preferably controllable by the control device (13).Cavity inspection device according to one of claims 1 to 3, characterised in that the control device (13) is designed in such a way that it takes into account movement information about a movement, in particular intended movement of the camera (10, 12), when controlling the exposure settings and / or the illumination device.Cavity inspection device according to any one of the preceding claims, characterized in that the control device (13) is configured to control the exposure settings and / or the illumination device (14, 16) on the basis of a recognized surface property which, according to taken into account movement information in a subsequent camera position, will be essential for the image exposure.Cavity inspection device according to one of the preceding claims, characterized in that the control device (13) and the illumination device (14, 16) are configured in such a way that the brightness of the illumination device (14, 16) and / or the light temperature and / or the emission characteristic of the illumination device (14, 16) can be changed.Cavity inspection device according to one of the preceding claims, characterized in that the control device (13) and the camera (10, 12) are configured in such a way that the aperture and / or shutter time of the camera (10, 12) can be changed by the control device.Cavity inspection device according to one of the preceding claims, characterized in that the surface property defines a reflection property or that the control device (13) is designed such that it assigns reflection properties or illumination settings to detected surface properties, for example on the basis of reflection properties and / or illumination settings stored in a table.Cavity inspection device according to one of the preceding claims, characterized in that the control device (13) is designed in such a way that it determines and preferably stores reflection properties and / or illumination settings associated with unknown surface properties by measurement.Cavity inspection device according to one of the preceding claims, characterized bytwo or more cameras (10, 12) connected to the control device (13) and illumination devices (14, 16) associated with the cameras (10, 12), which are controllable by the control device (13), wherein the control device (13) is designed such that the image analysis device recognizes surface properties in at least one image of at least one first camera (10), and that the control device (13) controls the exposure settings and / or the illumination device (14, 16) of at least one second camera (12) on the basis of these recognized surface properties.Method for controlling the illumination of a cavity inspection device, comprising the following steps: - capturing a camera image - analyzing the camera image with regard to the surface properties which have an influence on the reflection and illumination in individual image regions - analyzing whether, in a next step, a movement of an image section towards a region with changed surface properties is to take place - adapting the exposure and / or illumination to these changed surface properties before or during the movement of the image section towards the region with changed surface properties.Method according to claim 11, characterised bythe following additional steps: - after the movement has taken place, checking whether the selected exposure settings and / or illumination settings are correct, - adapting the exposure and / or illumination if the exposure settings and / or illumination settings are not correct, - optionally correcting the exposure and / or illumination settings assigned to the detected surface properties on the basis of the adaptation that has taken place.

Images