Abnormality determination device, abnormality determination system, and abnormality determination method

Abstract

To relatively easily and automatically detect that imaging with an in-vehicle camera is interrupted.SOLUTION: An abnormality determination device comprises an interruption determination unit (211) which uses the number of edges extracted by image processing of an image captured with a camera (10) and a gain level value obtained by digitizing a gain volume in capturing the image and determines that imaging with the camera (10) is interrupted, in a case where a first condition that the gain level value is within a normal range and the number of edges is less than a prescribed first setting value is satisfied.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an abnormality determination device that determines whether or not an abnormality has occurred in a camera. [Background technology]

[0002] 2. Description of the Related Art Conventionally, there is known a surveillance camera device that detects that a person is masked when packing tape or the like is attached to the surveillance camera (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2003-134504 A Summary of the Invention [Problem to be solved by the invention]

[0004] In recent years, there has been a demand for installing in-car cameras inside railroad cars. In order to deal with situations where the imaging by the in-car camera is obstructed, a system that can automatically detect the obstruction of imaging by the in-car camera is required. The technology described in Patent Document 1 does not assume the use of an in-car camera, and there is room for improvement. [Means for solving the problem]

[0005] In order to solve the above problems, an abnormality judgment device according to one embodiment of the present invention is an abnormality judgment device that judges whether or not an abnormality has occurred in an in-car camera that captures images inside a railway vehicle, and is equipped with an interference judgment unit that judges that imaging by the in-car camera is being interfered with when a first condition is met that the gain level value is within a normal range and the number of edges is below a predetermined first set value, using (i) a number of edges extracted by performing image processing on an image captured by the in-car camera, and (ii) a gain level value that quantifies a gain amount by which an electrical signal generated in the imaging element of the in-car camera when capturing the image is amplified to a predetermined level. Effect of the Invention

[0006] According to one aspect of the present invention, it is possible to automatically detect with relative ease whether an object that transmits light relatively easily is covering the lens of the in-vehicle camera and thus interfering with imaging by the in-vehicle camera. [Brief description of the drawings]

[0007] [Figure 1] 1 is a block diagram showing a configuration of an abnormality determination system according to a first embodiment of the present invention. [Diagram 2] 5 is a flowchart showing an example of a flow of processes executed by the abnormality determination system in the first embodiment of the present invention. [Diagram 3] 6 is a flowchart showing an example of a flow of processes executed by the abnormality determination system in another configuration example of the first embodiment of the present invention. [Figure 4] 6 is a flowchart showing an example of a flow of processes executed by the abnormality determination system in another configuration example of the first embodiment of the present invention. [Diagram 5] FIG. 11 is a block diagram showing the configuration of an abnormality determination system according to a second embodiment of the present invention. [Figure 6] 10 is a flowchart showing an example of a flow of processes executed by an abnormality determination system according to a second embodiment of the present invention. [Figure 7] FIG. 11 is a block diagram showing the configuration of an abnormality determination system according to a third embodiment of the present invention. [Figure 8] 11 is a flowchart showing an example of a flow of processes executed by an abnormality determination system according to a third embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] [Embodiment 1] An embodiment of the present invention will be described in detail below. One aspect of the present invention relates to an abnormality determination device that determines whether or not imaging by an in-car camera (on-board camera) installed in a railway vehicle is obstructed. Hereinafter, a state in which imaging by the in-car camera is not obstructed may be referred to as a "normal state." Furthermore, a railway vehicle may be simply abbreviated to a vehicle.

[0009] <System configuration> The configuration of the abnormality determination system in this embodiment will be described below with reference to Fig. 1. Fig. 1 is a block diagram showing the configuration of the abnormality determination system in this embodiment.

[0010] As shown in Fig. 1, the abnormality determination system 100 in this embodiment includes a camera (in-vehicle camera) 10 installed in a vehicle, and a management device (controller) 20 having a function as an abnormality determination device. The camera 10 includes a lens 11, an aperture (aperture diaphragm) 12, an imaging element (image sensor) 13, a gain adjustment unit 14, an A / D converter (analog-to-digital conversion circuit) 15, and an image processing unit 16. The image processing unit 16 includes an image generation unit 161 and an edge number extraction unit 162. The management device 20 includes a control unit 21, a communication unit 22, and a storage unit 23. The control unit 21 includes an obstruction determination unit 211.

[0011] (camera) The camera 10 is an imaging device for capturing images of passengers and on-board facilities in a vehicle. The camera 10 is installed inconspicuously in the vehicle. Examples of places where the camera 10 is installed include narrow spaces above each door, the ceiling, lighting areas, etc.

[0012] In the camera 10, light from a subject is captured by the image sensor 13 through the lens (front lens) 11 and the aperture 12 and converted into an analog electrical signal (luminance signal). For example, a CCD (Charged-Coupled Device) sensor, a CMOS (Complementary Metal-Oxide Semiconductor) sensor, a MOS (Metal-Oxide Semiconductor) sensor, etc. can be used as the image sensor 13. Regardless of the type of image sensor used in the image sensor 13, a gain adjustment unit 14 performs gain control on the electrical signal output from the image sensor 13.

[0013] The camera 10 in this embodiment is a compact camera having an imaging section with a fixed focus and a fixed aperture. The focal length of the camera 10 may be 3 mm or less, for example, 1.1 mm to 2.7 mm. The aperture value (F value) of the camera 10 may be 4 or less, for example, 2.0 to 2.8. The camera 10 may be a wide-angle camera having a wide imaging field of view, for example, a horizontal angle of view of 120 degrees to 230 degrees and a vertical angle of view of 80 degrees to 170 degrees. By using a wide-angle camera, the number of cameras 10 installed in the vehicle can be reduced. For example, the cameras 10 may be installed in a staggered arrangement (alternate arrangement on the left and right) at four locations in the vehicle. The camera 10 may have a lens aperture of the lens 11 of 6 to 15 mm, and the sensor size of the imaging element 13 may be, for example, 1 / 3 inch or 1 / 2.8 inch.

[0014] In order to construct the abnormality determination system 100, for example, when installing an additional camera 10 in an existing vehicle, using an existing in-vehicle camera, or building a new vehicle incorporating the camera 10, the following is required. That is, in reality, it is required to reduce the cost of constructing the abnormality determination system 100, as well as to reduce the cost of the camera 10 and to make the space required for installation compact. In response to these demands, the abnormality determination system 100 of the present embodiment can use a relatively inexpensive and small wide-angle camera as the camera 10, and can be relatively easily adapted.

[0015] Since the analog electrical signal generated by the imaging element 13 is weak, it is amplified by the gain adjustment unit 14. The gain adjustment unit 14 has a function known as AGC (Auto Gain Control). The AGC is a function for stabilizing the electrical output signal level as follows. For example, when imaging a dark subject, the amount of light taken into the imaging element 13 is relatively small, so the AGC increases the gain amount (amplification degree) so that the image becomes brighter. Conversely, when the amount of light taken into the imaging element 13 is relatively large, the AGC decreases the gain amount. This makes it possible to keep the electrical signal level output from the gain adjustment unit 14 constant.

[0016] The analog electrical signal amplified by the gain adjustment unit 14 is transmitted to the A / D converter 15. In the A / D converter 15, the analog electrical signal is converted into a digital signal (A / D conversion). Thereafter, the image generation unit 161 of the image processing unit 16 generates image data based on the digital signal received from the A / D converter 15. Hereinafter, the image data generated by the image generation unit 161 of the image processing unit 16 may be referred to as a captured image.

[0017] Like a typical in-car camera, camera 10 may have resistance to vibration and resistance to voltage fluctuations in power supply from the vehicle power supply. Camera 10 may have performance that complies with standards related to railway vehicles, for example, Japanese standards (e.g., JIS E5006) or international standards (e.g., EN 50155). Publicly known components may be used for camera 10. In this specification, detailed descriptions of the processes performed by publicly known components included in abnormality determination system 100 will be omitted as appropriate to avoid redundancy.

[0018] The camera 10 can continuously perform the above-mentioned series of processes. In this embodiment, an example will be described in which the camera 10 generates moving image data using a plurality of image data arranged in time series, but the camera 10 may generate image data of still images at a predetermined time interval.

[0019] Generally, for a vehicle in operation (operating for business), there are regulations (see, for example, JIS E4016:2009) regarding the illuminance inside the vehicle, and the vehicle is designed to have an in-vehicle environment (lighting, etc.) that satisfies the regulations while in operation. Therefore, the space inside the vehicle in operation that the camera 10 is to capture is maintained at a brightness (for example, illuminance) greater than a certain level regardless of whether it is day or night. As a result, when the vehicle is in operation and the camera 10 is in a normal state, the gain amount in the gain adjustment unit 14 varies within a range smaller than the maximum gain amount, with the gain amount being the maximum when the space inside the vehicle is at the minimum brightness of the specified range. For convenience of explanation, this range is referred to as the "gain variation range during normal operation." In addition, when the camera 10 captures the space inside the vehicle in operation with a relatively wide imaging field of view (angle of view), the captured image may include multiple images of artificial objects and people (passengers) with vertical edges.

[0020] For example, when the lens 11 is covered by an object that transmits light relatively little, the gain amount in the gain adjustment unit 14 may exceed the gain variation range during normal operation, and based on this, it can be determined that imaging by the camera 10 is being obstructed. On the other hand, when the lens 11 is covered by an object that transmits light relatively easily, the amount of light taken into the imaging element 13 does not change (does not decrease significantly), so the gain amount in the gain adjustment unit 14 may be within the gain variation range during normal operation. Even when the gain amount in the gain adjustment unit 14 is within the gain variation range during normal operation, but imaging by the camera 10 is actually being obstructed, a system is required that can automatically detect imaging obstruction.

[0021] The inventors of the present invention have conducted intensive research and come up with the following new idea. That is, in an imaging environment of an in-vehicle camera, at least the in-vehicle equipment is included in the imaging field of the installed in-vehicle camera. For the variation range of the number of edges extracted from an image captured by the camera 10 in a normal state while the vehicle is running, the lower limit of the variation range can be set (specified) in advance based on the number of edges extracted from the image of the in-vehicle equipment included in the captured image. Then, in order to determine whether or not the imaging of the camera 10 is obstructed despite being within the gain variation range during normal operation, the number of edges extracted from the captured image can be used as a determination index. In other words, when the number of edges extracted from the captured image is below the lower limit of the preset variation range of the number of edges, it can be determined that the imaging of the camera 10 is obstructed. Since such arithmetic processing can be performed relatively simply, there are relatively few requirements for the performance of the arithmetic circuit. In addition, for example, data obtained by an illuminance sensor, an infrared receiving sensor, etc. is not required, and various components are not required to be added to the camera or the system. These are practically great advantages in implementing the abnormality determination system 100 in a vehicle.

[0022] In the abnormality determination system 100 in this embodiment, the gain amount of the electric signal generated by the image sensor 13 is quantified in the gain adjustment unit 14 in order to amplify the electric signal to a predetermined level. In this specification, this quantified value is referred to as a gain level value. A specific calculation method of the gain level value is not particularly limited, and a predetermined rule can be set in advance so that the gain level value is quantified according to the predetermined rule. For example, the gain level value is a value obtained by quantifying the amplification rate, and may be a value within a range of 0 to 255 (256 stages). The predetermined level related to the amplification of the electric signal in the gain adjustment unit 14 is set in advance, and the gain variation range during the normal operation can be set so as to correspond to the predetermined level. Then, the normal range of the gain level value can be specified based on the gain variation range during the normal operation. The maximum gain amount in the gain variation range during the normal operation corresponds to the upper limit of the normal range of the gain level value. The lower limit of the normal range of the gain level value may or may not be set. The normal range of the gain level value may be set appropriately in consideration of the imaging conditions such as the shutter speed of the camera 10.

[0023] The functional block and timing for specifying the gain level value are not particularly limited. For example, the gain level value may be specified by the gain adjustment unit 14 based on the gain amount in the gain adjustment unit 14. In this case, the analog electrical signal amplified by the gain adjustment unit 14 at a certain gain amount and the information on the gain level value based on the gain amount may be linked to each other. Alternatively, the gain level value may be specified by, for example, the image processing unit 16. In this case, the image processing unit 16 acquires information on the gain amount from the gain adjustment unit 14. Then, the data may be transmitted from the camera 10 to the management device 20 by linking the gain level value corresponding to the gain amount in the gain adjustment unit 14 in a series of processes until a certain captured image (for convenience of explanation, referred to as the captured image X1) is generated by the image generation unit 161 of the image processing unit 16 and the captured image X1.

[0024] The image processing unit 16 may perform known pre-processing on the captured image generated by the image generating unit 161. The pre-processing is not particularly limited, but may be, for example, a filter process for removing noise. The filter process may be, for example, a median filter, a mean filter, or a Gaussian filter.

[0025] The edge number extraction unit 162 performs a process of extracting the number of edges from the captured image after preprocessing. In the abnormality determination system 100 in this embodiment, the number of edges is the number of vertical edges extracted by performing image processing on the captured image. The edge number extraction unit 162 performs a process of extracting the number of vertical edges (the number of vertical edges) from the captured image after preprocessing. This makes it easier to reduce the amount of calculation processing compared to the case of extracting the number of edges in both the vertical and horizontal directions. A known method can be used as the process of extracting the number of vertical edges. For example, the edge number extraction unit 162 extracts vertical edges (vertical edges) from the captured image after preprocessing (image data after noise removal) using the following Sobel filter (vertical edge extraction filter) Gy.

[0026]

number

[0027] Then, the edge number extraction unit 162 performs binarization processing on the image data from which the vertical edges have been extracted, thereby generating a vertical edge binary image. In the binarization processing, for example, if the luminance value of a certain pixel is equal to or greater than a threshold, the value of the pixel is converted to 255, and if the luminance value is less than the threshold, the value of the pixel is converted to 0. Next, the number of pixels having a luminance value of 255 in the vertical edge binary image is counted. This makes it possible to extract the number of vertical edges from the captured image. Alternatively, instead of the binarization processing, the number of vertical edges may be extracted by counting the number of pixels having a luminance value equal to or greater than a threshold. The number of vertical edges extracted by the edge number extraction unit 162 is linked to the captured image from which the number of vertical edges was extracted.

[0028] When the vehicle is in operation and the camera 10 is in a normal state, the fluctuation range of the edge number extracted by the edge number extraction unit 162 is referred to as the "edge number fluctuation range during normal operation" for ease of explanation. The lower limit (a predetermined first set value) of the edge number fluctuation range during normal operation can be set in advance based on the number of edges extracted from an image captured by the camera 10 installed inside the vehicle with no passengers included in the imaging field of view.

[0029] In the abnormality determination system 100 of this embodiment, data on the gain level value and the number of vertical edges linked to the image data are transmitted from the camera 10 to the management device 20 together with image data (digital video signal). The specific means for transmitting data from the camera 10 to the management device 20 is not particularly limited, and any known transmission means can be used.

[0030] In another embodiment of the present invention, the edge number extraction unit 162 may be provided in the camera 10 as a calculation circuit different from the image processing unit 16 including the image generation unit 161. Also, an external device different from the camera 10 may have the edge number extraction unit 162, in which case the number of edges may be extracted by transferring the image data to the external device. In the abnormality determination system 100, it is only necessary to be able to extract the number of edges from the captured image, and there is no particular limitation on the device in which the edge number extraction unit 162 is provided.

[0031] (Management device) The details of the processing of each part in the management device 20 will be described later together with the processing flow of the system. In general, the control unit 21 has a processor such as a CPU (central processing unit) and a memory, and the interference determination unit 211 of the control unit 21 determines whether or not the imaging by the camera 10 is being interfered with, based on the above-mentioned gain level value and edge number information.

[0032] The communication unit 22 communicates with the camera 10 via a known transmission means, and also communicates with the outside of the device via the network 30. The communication unit 22 may communicate with, for example, a crew room inside the train, or may communicate with a train operation control center outside the train. The camera 10 and the management device 20 are connected, for example, by a vehicle network. The network 30 may be, for example, a vehicle network, and when it is determined that the imaging by the camera 10 is being disturbed, the communication unit 22 may notify the crew of that fact. Alternatively, the network 30 may be, for example, a wireless communication network, and when it is determined that the imaging by the camera 10 is being disturbed, the communication unit 22 may notify an external person to be notified of that fact. In this way, the communication unit 22 can function as a notification unit that notifies an abnormality in the camera 10 when it is detected.

[0033] The storage unit 23 stores various data including image data, and gain level values ​​and edge counts associated with the image data. The storage unit 23 may store these various data in a ring buffer. The specific form of the storage unit 23 is not particularly limited, and may have, for example, a volatile memory or a non-volatile memory. The program executed in the control unit 21 may be stored in the storage unit 23, for example. The storage unit 23 may also store information regarding the normal range of the gain level value and the edge count fluctuation range during normal operation.

[0034] <Processing flow> Next, the flow of processing (abnormality determination processing) executed by the abnormality determination system 100 will be described below with reference to FIG. 2. FIG. 2 is a flowchart showing an example of the flow of processing executed by the abnormality determination system 100 in this embodiment. Note that the abnormality determination processing described below is executed on the premise that the vehicle is in operation. The abnormality determination system 100 executes the abnormality determination processing when, for example, a signal indicating that the vehicle is in operation is received from a vehicle information control device. In this specification, a state in which a vehicle is in commercial operation is simply referred to as "in operation". The state in which a vehicle is in operation does not include, for example, a state in which the vehicle is in a deadheading operation.

[0035] As shown in FIG. 2, first, a captured image is generated in the camera 10 (step S11). As described above, the image sensor 13 receives light from a subject via the lens 11 and the aperture 12 to generate an analog electrical signal, and the captured image is generated through processing in the gain adjustment unit 14, the A / D converter 15, and the image processing unit 16. Based on the gain amount in the gain adjustment unit 14, the gain adjustment unit 14 specifies a gain level value according to, for example, a preset rule (step S12). Step S12 may be performed before the completion of step S11, or may be performed in parallel with step S11. In addition, the edge number extraction unit 162 extracts the number of vertical edges from the captured image (step S13). Steps S12 and S13 may be performed in reverse order or in parallel.

[0036] The management device 20 acquires various data transmitted from the camera 10 (step S14). The various data includes the captured image and information on the gain level value and the number of edges associated with the captured image.

[0037] If the gain level value is within the normal range (YES in step S15) and the number of edges is below a predetermined first set value (YES in step S16), the interference determination unit 211 increments the counter value of the imaging error counter (step S18). The normal range of the gain level value can be set in advance based on the gain fluctuation range during normal operation as described above. Also, the predetermined first set value related to the number of edges can be set in advance as the lower limit of the edge number fluctuation range during normal operation as described above. The imaging error counter may be temporarily formed in a memory possessed by the control unit 21, or may be included in the storage unit 23.

[0038] If the counter value of the imaging error counter does not exceed the predetermined judgment threshold (NO in step S19), the process returns to step S11, and the abnormality judgment process from step S11 is performed again. The predetermined judgment threshold for the imaging error counter can be set appropriately, and the specific value is not particularly limited. If the counter value of the imaging error counter exceeds the predetermined judgment threshold (YES in step S19), the interference judgment unit 211 judges that imaging by the camera 10 is being interfered with (step S20). If the interference judgment unit 211 judges that imaging by the camera 10 is being interfered with, it notifies the person to be notified of this fact using the communication unit 22, for example, via the network 30.

[0039] Furthermore, if the result of step S15 is YES and the result of step S16 is NO, the interference determination unit 211 clears the counter value of the imaging error counter (step S17). If the vehicle is in operation and imaging by the camera 10 continues after step S17, the abnormality determination process shown in FIG. 2 is performed again.

[0040] In this way, the obstruction determination unit 211 uses (i) the number of edges extracted by performing image processing on the captured image captured by the camera 10, and (ii) a gain level value that quantifies the gain amount by amplifying to a predetermined level the electrical signal generated in the imaging element 13 of the camera 10 when capturing the captured image. Then, the obstruction determination unit 211 determines that imaging by the camera 10 is being obstructed when a first condition is satisfied, that is, the gain level value is within a normal range and the number of edges is below a predetermined first set value. Unlike the example shown in Fig. 2, steps S18 and S19 may not be included in the processing flow of the abnormality determination system 100.

[0041] According to management device 20 of this embodiment, for example, when lens 11 of camera 10 is covered with something that transmits light relatively easily (such as translucent tape), it is possible to automatically detect, relatively easily, that imaging by camera 10 is being hindered based on the satisfaction of the first condition. This effect can also be obtained if the front surface of lens 11 is partially covered with something that transmits light relatively easily.

[0042] Furthermore, the present invention is not limited to the example of attaching an object that is relatively light-transmitting to the lens 11, and the management device 20 can relatively easily automatically detect various types of mischief (fraudulent acts) that interfere with imaging by the camera 10, such as acts that reduce the number of edges extracted from the captured image, thereby interfering with imaging by the camera 10.

[0043] In the management device 20 of this embodiment, the interference determination unit 211 determines that the imaging by the camera 10 is being interfered with when the state in which the first condition is satisfied continues for a predetermined time. This predetermined time is determined corresponding to the execution interval of the abnormality determination process in the example shown in FIG. 2 and a predetermined determination threshold for the imaging error counter. In reality, even if no tampering has been performed on the lens 11 of the camera 10, the number of edges may suddenly change for a short time due to some factor. In the example shown in FIG. 2, the interference determination unit 211 can reduce the possibility of determining that the imaging by the camera 10 is being interfered with based on such an abrupt short-term change in the number of edges. This can further improve the accuracy of the determination by the interference determination unit 211.

[0044] Also, even if the gain level value is not within the normal range (NO in step S15 above), the interference determination unit 211 increments the counter value of the imaging error counter (step S18). In this case, regardless of whether the first condition is satisfied or not, interference determination unit 211 can determine that imaging by camera 10 is being interfered with based on the fact that the gain condition that the gain level value is within the normal range is not satisfied. For example, when the lens is covered with something that is relatively poorly transparent to light (such as packing tape), it can be determined that imaging is being interfered with. This effect can also be obtained if the front surface of lens 11 is partially covered with something that is relatively poorly transparent to light.

[0045] In the abnormality determination system 100 of this embodiment, in order to determine whether or not imaging by the camera 10 is being obstructed, it is not necessary to perform relatively advanced arithmetic processing in the camera 10 and the management device 20, and it is also not necessary to add various components to the camera 10. By operating the camera 10 using a relatively low-performance arithmetic circuit, it is easy to reduce the amount of heat generated by the operation of the camera 10. As a result, it is easy to reduce the introduction cost and running cost of the abnormality determination device and abnormality determination system that determine whether or not an abnormality has occurred in the camera 10 installed in a vehicle.

[0046] (Abnormality determination method) The scope of the present invention also includes an abnormality determination method including the processing performed by the management device 20 of the abnormality determination system 100 in this embodiment. This also applies to the abnormality determination systems 100 and management devices 20 in each embodiment and configuration example described in this specification, and repeated explanation will be omitted, but the abnormality determination method corresponding to the processing of the abnormality determination system 100 and management device 20 described below also falls within the scope of the present invention.

[0047] An abnormality determination method in one aspect of the present invention is an abnormality determination method for determining whether or not an abnormality has occurred in an in-car camera that captures images inside a railway vehicle, and includes an interference determination step of determining that imaging by the in-car camera is being interfered with if a first condition is met that the gain level value is within a normal range and the number of edges is below a predetermined first set value, using (i) a number of edges extracted by performing image processing on an image captured by the in-car camera, and (ii) a gain level value that quantifies a gain amount by which an electrical signal generated in an imaging element of the in-car camera when capturing the image is amplified to a predetermined level.

[0048] <Another configuration example> (a) In the anomaly determination system 100 according to one embodiment of the present invention, the camera 10 does not need to have a fixed focus and a fixed aperture. The camera 10 may have, for example, a varifocal lens. Various setting values ​​used in the anomaly determination process may be preset so as to correspond to the focal length and aperture values ​​identified in the process of generating a captured image.

[0049] (b) The management device 20 in one embodiment of the present invention may be installed as a part of a general vehicle information control device, and in this case, each part of the existing vehicle information control device may be used to function as the abnormality determination device. Alternatively, the management device 20 may be installed as a device separate from the vehicle information control device.

[0050] (c) In the example shown in FIG. 1, the management device 20 has a function as an abnormality determination device, but in one embodiment of the present invention, for example, the camera 10 may have a function as an abnormality determination device, and in this case, the camera 10 may have the interference determination unit 211. Alternatively, in another embodiment of the present invention, an external device located at a remote location and the camera 10 may be communicably connected through a network, and the external device may have a function as an abnormality determination device. In addition, in the abnormality determination system 100 in one embodiment of the present invention, the gain level value may be specified by the management device 20, and in this case, the captured image and the gain amount are linked, and data is transmitted from the camera 10 to the management device 20. The camera 10 and the management device 20 may be connected to each other by wireless communication.

[0051] (d) In the camera 10 according to an embodiment of the present invention, the edge number extraction unit 162 may extract the number of edges in both the vertical and horizontal directions from the captured image, and a known method such as the Canny method may be used as a specific process. Compared to extracting only vertical edges, the amount of calculation processing increases when the Canny method is used, but edges that form the contours of images of artifacts and people (passengers) included in the captured image can be extracted relatively precisely. A detailed description of the Canny method is omitted.

[0052] (e) Fig. 3 is a flowchart showing an example of the flow of processing executed by the abnormality determination system 100 in another configuration example of an embodiment of the present invention. The example shown in Fig. 3 differs from the example shown in Fig. 2 described above in that the lower limit (value that becomes the lower limit) of the normal range of the gain level value is set, and a second set value higher than the upper limit of the normal range of the gain level value, which is applied according to the operating state (mode) of the vehicle, is set in advance. The differences from the example shown in Fig. 2 will be described below, and the description of the steps described above with reference to Fig. 2 will not be repeated. When the second condition that the gain level value is higher than the second set value is satisfied, the interference determination unit 211 determines that the imaging by the camera 10 is being interfered with, regardless of whether the first condition is satisfied or not.

[0053] The lower limit of the normal range of the gain level value (lower limit value) may be appropriately set according to the expected image capture interference. For example, when high-intensity light (high-intensity LED light, laser light, etc.) is irradiated to the camera 10, the gain level value may be extremely low. In the example shown in FIG. 3, two modes may be set in advance for the vehicle operation state. For example, in a night train, the illuminance inside the vehicle may be relatively dark at night, and the vehicle operation state in this case may be set as a dark mode (second mode), and a state other than the dark mode, i.e., a state during normal operation, may be set as a light mode (first mode). The vehicle operation state at the time (current time) when the abnormality determination system 100 executes the process may be specified based on, for example, a clock (real-time clock) built into the management device. The vehicle operation state may also be specified based on lighting control information acquired via a network.

[0054] 3, specifically, the above-mentioned processes from steps S11 to S14 are performed first. Next, if the gain level value is not within the normal range (NO in step S15), the interference determination unit 211 determines whether or not the gain level value is below the lower limit of the normal range (step S21). If YES in step S21, the process proceeds to step S18. This is because there is a possibility that an act of interference with imaging of the camera 10 is being performed, or that an abnormality has occurred in imaging of the camera 10 due to some factor.

[0055] If the result is NO in step S21, that is, if the result is YES in step S15 and NO in step S21, and therefore the gain level value is above the upper limit of the normal range, the interference determination unit 211 determines whether the current operation state is the dark mode or not (step S22). Note that the timing at which the mode identification process for identifying the current operation state is performed and the functional block for performing the mode identification process are not particularly limited.

[0056] If NO in step S22, when the vehicle is not in the dark mode (i.e., in the light mode), and NO in step S15 and NO in step S21, that is, the gain level value exceeds the upper limit of the normal range, the process proceeds to step S18. If YES in step S22, the interference determination unit 211 determines whether the gain level value exceeds a predetermined second set value (step S23). If YES in step S23, the process proceeds to step S18. If NO in step S23, the process proceeds to step S16. The significance (effect) of such a processing flow will be described below.

[0057] First, it can be said that the upper limit of the gain variation range in the dark mode is higher than that of the gain variation range in the light mode (normal operation). Providing a second setting value to be applied in the dark mode means expanding the upper limit of the gain level value to correspond to the gain variation range in the dark mode.

[0058] In the example shown in Fig. 3, by including steps S22 and S23, the following case can be appropriately handled. For example, the lens 11 may be partially covered by an object that is relatively poorly transparent to light, causing the gain level value to exceed the upper limit of the normal range but not exceed the second set value. In this case, in the example shown in Fig. 3, if the vehicle is operating in the bright mode, NO is determined in step S15, NO in step S21, and NO in step S22, and the process proceeds to step S18 (imaging interference can be detected).

[0059] On the other hand, when the vehicle is in the dark mode, the illuminance inside the vehicle is relatively dark. Therefore, for example, in the dark mode, even though the lens 11 is not obstructed from imaging, the gain level value may exceed the upper limit of the normal range and may not exceed the second set value. In this case, the process proceeds to step S16 by determining NO in step S15, NO in step S21, YES in step S22, and NO in step S23. This means that the upper limit for the dark mode is within the range of the virtual normal range of the expanded gain level value (virtually, YES in step S15 in the dark mode). And, for example, when the lens 11 is partially covered by an object that is relatively difficult to transmit light, even if the gain level value in the light mode exceeds the upper limit of the normal range and does not exceed the second set value, the gain increases overall in the dark mode, so that the gain level value exceeds the second set value. As a result, the process proceeds to step S18 by determining NO in step S15, NO in step S21, YES in step S22, and YES in step S23 (imaging obstruction can be detected).

[0060] In the example shown in FIG. 3, the interference determination unit 211 determines that the imaging by the camera 10 is being interfered with when the state in which the second condition that the gain level value exceeds the second set value continues for a predetermined time in the dark mode. This predetermined time is determined corresponding to the execution interval of the abnormality determination process in the example shown in FIG. 3 and a predetermined determination threshold value for the imaging error counter. In reality, even if no tampering has been performed on the lens 11 of the camera 10, the gain level value may suddenly change for a short time due to some factor. In the example shown in FIG. 3, the interference determination unit 211 does not determine that the imaging by the camera 10 is being interfered with based on such an abrupt short-time change in the gain level value. This can further improve the accuracy of the determination by the interference determination unit 211.

[0061] (f) Fig. 4 is a flowchart showing an example of the flow of processing executed by the abnormality determination system 100 in another configuration example of one embodiment of the present invention. The example shown in Fig. 4 differs from the example shown in Fig. 3 described above in that when the gain level value is below the lower limit of the normal range, a determination process based on the number of edges is further performed. Only the differences from the example shown in Fig. 3 will be described below.

[0062] 4, specifically, when the gain level value is below the lower limit of the normal range (YES in step S21), the interference determination unit 211 determines whether the number of edges is below a predetermined first set value (step S24). If NO in step S24, the process proceeds to step S18, and if YES in step S24, the process returns to step S11 and performs the abnormality determination process from step S11 again.

[0063] The above means that when the gain level value is below the lower limit of the normal range and the number of edges is equal to or less than a predetermined number, it is not determined that imaging by camera 10 is being obstructed. For example, when flare occurs due to the ambient light in the imaging field of view of camera 10 being too strong (the imaging field of view is too bright), the number of edges may decrease. In such a state, in the example shown in FIG. 4, it is possible to effectively reduce the possibility that the obstruction determination unit 211 will erroneously determine that imaging by camera 10 is being obstructed.

[0064] (g) In the management device 20 according to an embodiment of the present invention, the imaging error counter may be configured to be decremented in step S18. In step S19, the interference determination unit 211 can determine that imaging by the camera 10 is being interfered with when the counter value of the imaging error counter falls below a predetermined determination threshold. The interference determination unit 211 only needs to be able to determine the presence or absence of imaging interference based on whether or not the state in which the first condition is satisfied has continued for a predetermined period of time, and a specific method for doing so is not particularly limited. This is the same in each embodiment described below, and a repeated description will be omitted.

[0065] (h) In the abnormality determination system 100 according to one embodiment of the present invention, two types of imaging error counters (a first imaging error counter and a second imaging error counter) may be used for determination in the interference determination section 211. Specifically, the following may be used.

[0066] That is, the interference determination unit 211 increments the counter value of the first imaging error counter when the gain level value is within the normal range and the number of edges is below the predetermined first set value. When the counter value of the first imaging error counter exceeds the predetermined first determination threshold, the interference determination unit 211 determines that imaging by the camera 10 is being interfered with. Furthermore, when the gain level value is not within the normal range and exceeds the predetermined second set value, the interference determination unit 211 increments the counter value of the second imaging error counter. When the counter value of the second imaging error counter exceeds the predetermined second determination threshold, the interference determination unit 211 determines that imaging by the camera 10 is being interfered with. When the gain level value is within the normal range and the number of edges is not below the predetermined first set value, the interference determination unit 211 clears the counter values ​​of the first imaging error counter and the second imaging error counter.

[0067] The first imaging error counter and the second imaging error counter may be temporarily formed in a memory of the control unit 21, or may be included in the storage unit 23. The predetermined first judgment threshold and the predetermined second judgment threshold are set in advance. The predetermined first judgment threshold and the predetermined second judgment threshold may have different values.

[0068] The management device 20 can distinguish between a case where the state where the first condition is satisfied continues and a case where the state where the second condition is satisfied continues by judging the imaging interference based on the counter values ​​of the first imaging error counter and the second imaging error counter. Therefore, the accuracy of the judgment by the interference judgment unit 211 can be further improved. As described above, in the dark mode, the gain level value may exceed the upper limit of the normal range and may not exceed the second set value. In this case, for example, due to fluctuations in the gain level value (noise, etc.), the judgment results in steps S15 and S16 may not be stable. Even in such a case, by using different imaging error counters, even if the interference judgment unit 211 erroneously judges that the first condition or the second condition is satisfied, the imaging error counter can be reset in step S17 before the interference judgment unit 211 judges that the state where the first condition is satisfied continues or the state where the second condition is satisfied continues. As a result, the possibility that the interference judgment unit 211 erroneously judges that the imaging by the camera 10 is interfered with can be reduced.

[0069] Similarly to the above, a third imaging error counter may be used for the judgment in the interference judgment unit 211. In this case, when the gain level value is below the lower limit of the normal range, the interference judgment unit 211 increments the counter value of the third imaging error counter. When the counter value of the third imaging error counter exceeds a predetermined judgment threshold, the interference judgment unit 211 judges that imaging by the camera 10 is being interfered with.

[0070] [Embodiment 2] Another embodiment of the present invention will be described below with reference to Figures 5 and 6. Note that configurations other than those described in this embodiment are the same as those in the first embodiment.

[0071] In the first embodiment, an example of the abnormality determination system 100 including one camera 10 and the management device 20 has been described, but in the second embodiment, an example of the abnormality determination system 100 including a plurality of cameras 10 will be described. In the second embodiment, based on information from the plurality of cameras 10, the management device 20 determines whether or not the vehicle is in operation, and when the vehicle is in operation, performs the abnormality determination process described above.

[0072] Fig. 5 is a block diagram showing the configuration of the abnormality determination system in this embodiment. Fig. 6 is a flowchart showing an example of the flow of processing executed by the abnormality determination system in this embodiment. Note that the operation determination processing described below is executed on the premise that the vehicle has been started, i.e., that power is being supplied from a power source to each part of the abnormality determination system 100.

[0073] 5, the abnormality determination system 100 in this embodiment includes four cameras 10 that are installed in the same vehicle 40 and capture images of the interior of the vehicle. The four cameras 10 may be installed in the same vehicle 40 in a staggered arrangement (alternate left and right arrangement), for example. In the management device 20 of this embodiment, the control unit 21 includes an operation determination unit 212. The management device 20 may be installed in the vehicle 40 or outside the vehicle 40.

[0074] As shown in FIG. 6, first, the management device 20 acquires various data from each camera 10 (step S31). Next, the operation determination unit 212 judges whether or not all of the multiple gain level values ​​acquired from each camera 10 at the same time exceed a predetermined third set value (step S32). The operation determination unit 212 may perform the judgment in step S32 using data transmitted from each camera 10 at the same time, or may perform the judgment in step S32 based on a synchronization signal in the abnormality determination system 100. As long as the operation determination unit 212 can judge the operation state using data from each camera at the same time among the data acquired from each camera 10, the specific method is not particularly limited.

[0075] If the answer is YES in step S32, that is, if the lights inside the vehicle 40 are not on and the interior space is generally dark, the operation determination unit 212 determines that the vehicle is not in operation. In this case, the management device 20 does not perform abnormality determination processing by the interference determination unit 211. On the other hand, if the answer is NO in step S32, the operation determination unit 212 determines that the vehicle is in operation. In this case, the management device 20 controls the interference determination unit 211 to perform abnormality determination processing for each camera 10 (step S34).

[0076] The third set value is set to a value higher than a normal range of gain level values. The third set value may be set to a value higher than the above-mentioned second set value.

[0077] In the abnormality determination system 100 in this embodiment, it is not necessary to transmit a signal indicating whether or not the vehicle 40 is in operation to the management device 20, etc., via a vehicle network. This makes it easier to reduce system costs.

[0078] [Embodiment 3] Another embodiment of the present invention will be described below with reference to Figures 7 and 8. Note that configurations other than those described in this embodiment are the same as those in the first and second embodiments.

[0079] In the second embodiment, an example has been described in which it is determined whether or not a vehicle is in operation based on information from a plurality of cameras 10. The abnormality determination system 100 in the third embodiment can determine whether or not each of the cameras 10 has a malfunction based on information from the plurality of cameras 10.

[0080] Fig. 7 is a block diagram showing the configuration of the abnormality determination system in this embodiment. Fig. 8 is a flowchart showing an example of the flow of processing executed by the abnormality determination system in this embodiment.

[0081] 7, in the abnormality determination system 100 of this embodiment, the control unit 21 of the management device 20 includes a malfunction determination unit 213. Before the interference determination unit 211 makes a determination, the malfunction determination unit 213 compares a plurality of gain level values ​​acquired at the same time from each of a plurality of cameras 10 capturing an image of the interior of the same vehicle 40, and determines that a malfunction has occurred in a camera 10 having a gain level value whose difference from the median value of the plurality of gain level values ​​is greater than a predetermined fourth set value. The process executed by the abnormality determination system 100 is described in more detail below.

[0082] 8, first, the management device 20 acquires various data from each camera 10 (step S31). Next, the malfunction determination unit 213 compares multiple gain level values ​​acquired at the same time from each camera 10 capturing an image of the interior of the same vehicle 40 (step S41).

[0083] The malfunction determination unit 213 identifies the median of multiple gain level values ​​acquired from each camera 10 at the same time. If the number of cameras is an even number, the average value of two values ​​near the center can be used as the median. The malfunction determination unit 213 determines whether or not there is a camera 10 having a gain level value whose difference from the identified median value is greater than a predetermined fourth set value (step S42). If the answer is YES in step S42, the malfunction determination unit 213 increments the counter value of the malfunction counter of the corresponding camera 10 (step S43). The malfunction counter may be temporarily formed in a memory possessed by the control unit 21, or may be included in the storage unit 23.

[0084] If there is no camera 10 whose counter value of the malfunction counter exceeds the predetermined third judgment threshold (NO in step S44), the process returns to step S31, and the malfunction judgment process from step S31 is performed again. The above-mentioned predetermined third judgment threshold can be set appropriately, and the specific value is not particularly limited. If there is a camera 10 whose counter value of the malfunction counter exceeds the predetermined third judgment threshold (YES in step S44), the malfunction judgment unit 213 judges that a malfunction has occurred in the corresponding camera 10 (step S45). If the malfunction judgment unit 213 judges that a malfunction has occurred in the camera 10, the communication unit 22 uses the communication unit 22 to notify an external report target of that fact via, for example, the network 30.

[0085] Furthermore, if the result of step S42 is NO, the malfunction determination unit 213 clears the counter value of the malfunction counter of each camera 10 (step S46). In the abnormality determination system 100 of this embodiment, after step S46, the operation determination unit 212 determines the operating state of the vehicle (steps S32, 33). Then, in the management device 20 of this embodiment, if it is determined that the vehicle is in operation, the interference determination unit 211 performs an abnormality determination process for each camera 10 (step S34).

[0086] As described above, in management device 20 of this embodiment, before performing abnormality determination processing by obstruction determination section 211, malfunction determination processing is performed by malfunction determination section 213. This makes it possible to determine whether or not there is any mischief (fraudulent behavior) that obstructs image capture by camera 10 in vehicle 40, and also to determine whether or not there is a malfunction in each camera 10.

[0087] [Other configuration examples] In the anomaly determination system 100 in one embodiment, the edge number extraction unit 162 may extract the number of edges using data that is reduced in size compared to the image data generated by the image generation unit 161. In this case, the amount of calculation processing in the edge number extraction unit 162 can be effectively reduced.

[0088] Furthermore, in the anomaly determination system 100 in one embodiment, the edge number extraction unit 162 may divide the image data generated by the image generation unit 161 into a plurality of regions (for example, 2 to 8 regions) and extract the number of edges for each region. In this case, the obstruction determination unit 211 can perform an anomaly determination process for each region. This makes it easier to determine that imaging is being obstructed when the front surface of the lens 11 is partially covered by some kind of foreign matter.

[0089] Here, an image captured by the camera 10 having a wide imaging field of view may include areas of relatively low importance (non-attention areas). Examples of non-attention areas include areas in the captured image where lighting or a ceiling is captured, areas at the edge of the captured image that are captured by another camera 10 in the vehicle 40, and the like. Areas other than the non-attention areas are referred to as attention areas.

[0090] In an embodiment of the abnormality determination system 100, the attention area may be subject to the abnormality determination process, while the non-attention area may be excluded from the edge number extraction target by the edge number extraction unit 162 and from the abnormality determination process target by the obstruction determination unit 211. This can effectively reduce the amount of calculation processing in the edge number extraction unit 162. The predetermined first setting value in step S16 (see FIG. 2) may be set for the attention area.

[0091] In an embodiment of the abnormality determination system 100, the image data generated by the image generating unit 161 may be divided to include a plurality of attention areas, and a different first setting value may be set for each of the plurality of attention areas. The obstruction determination unit 211 can determine that imaging by the camera 10 is obstructed when at least one of the plurality of attention areas satisfies the above-mentioned first condition.

[0092] 〔summary〕 The abnormality determination device in aspect 1 of the present invention is an abnormality determination device that determines whether or not an abnormality has occurred in an in-car camera that captures images inside a railway vehicle, and includes an interference determination unit that determines that imaging by the in-car camera is being interfered with when a first condition is met that the gain level value is within a normal range and the number of edges is below a predetermined first set value, using (i) the number of edges extracted by performing image processing on the image captured by the in-car camera, and (ii) a gain level value that quantifies the gain amount by which an electrical signal generated in the imaging element of the in-car camera when capturing the image, is amplified to a predetermined level.

[0093] In the abnormality judgment device of aspect 2 of the present invention, in aspect 1, a predetermined second set value higher than the normal range of the gain level value is set, and when the second condition that the gain level value exceeds the second set value is satisfied, the interference judgment unit judges that imaging by the in-vehicle camera is being interfered with, regardless of whether the first condition is satisfied.

[0094] In the abnormality judgment device of aspect 3 of the present invention, in the above-mentioned aspect 1 or 2, the interference judgment unit judges that imaging by the in-vehicle camera is being interfered with when the state satisfying the first condition continues for a predetermined period of time.

[0095] In the abnormality judgment device of aspect 4 of the present invention, in aspect 2 or 3, the interference judgment unit judges that imaging by the in-vehicle camera is being interfered with when the state satisfying the second condition continues for a predetermined period of time.

[0096] In the abnormality judgment device of aspect 5 of the present invention, in any one of aspects 1 to 4, when all of the multiple gain level values ​​obtained at the same time from each of the multiple in-car cameras capturing images of the interior of the same railway vehicle exceed a predetermined third set value, the interference judgment unit does not judge whether imaging by the in-car cameras is being interfered with.

[0097] The abnormality judgment device in aspect 6 of the present invention, in any one of aspects 1 to 5, further includes a malfunction judgment unit that, before making a judgment by the interference judgment unit, compares the gain level values ​​obtained at the same time from each of the multiple in-vehicle cameras capturing images of the interior of the same railway vehicle, and judges that a malfunction has occurred in the in-vehicle camera having a gain level value whose difference from the median of the multiple gain level values ​​is greater than a predetermined fourth set value.

[0098] The abnormality determination device according to Aspect 7 of the present invention is any one of Aspects 1 to 6, wherein the number of edges is the number of vertical edges extracted by performing image processing on the captured image.

[0099] The abnormality determination system according to an eighth aspect of the present invention includes the abnormality determination device according to any one of the first to seventh aspects and the on-board camera installed in the railway vehicle.

[0100] The abnormality determination method in aspect 9 of the present invention is an abnormality determination method for determining whether or not an abnormality has occurred in an in-car camera that captures images inside a railway vehicle, and includes an interference determination step for determining that imaging by the in-car camera is being interfered with if a first condition is met that the gain level value is within a normal range and the number of edges is below a predetermined first set value, using (i) the number of edges extracted by performing image processing on the image captured by the in-car camera, and (ii) a gain level value that quantifies the gain amount by which the electrical signal generated in the imaging element of the in-car camera when capturing the image is amplified to a predetermined level, and the gain level value is within a normal range and the number of edges is below a predetermined first set value.

[0101] [Software implementation example] The functions of the management device 20 (hereinafter referred to as the "device") of the abnormality determination system 100 can be realized by a program for causing a computer to function as the device, and a program for causing a computer to function as each control block of the device (particularly each part included in the control unit 21).

[0102] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., a memory) as hardware for executing the program. The control device and storage device execute the program to realize each function described in each of the above embodiments.

[0103] The program may be non-transitory and may be recorded in one or more computer-readable recording media. The recording media may or may not be included in the device. In the latter case, the program may be provided to the device via any wired or wireless transmission medium.

[0104] In addition, some or all of the functions of each of the control blocks can be realized by a logic circuit. For example, an integrated circuit in which a logic circuit that functions as each of the control blocks is formed is also included in the scope of the present invention. In addition, the functions of each of the control blocks can be realized by, for example, a quantum computer.

[0105] Furthermore, each process described in each of the above embodiments may be executed by AI (Artificial Intelligence). In this case, the AI ​​may be executed by the control device or another device (for example, an edge computer or a cloud server).

[0106] [Additional notes] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention.

[0107] According to one aspect of the present invention, it is possible to automatically detect the interference with the imaging by the in-vehicle camera relatively easily, thereby contributing to the improvement of safety inside the vehicle. Such an effect also contributes to the achievement of, for example, Goal 11 of the Sustainable Development Goals (SDGs) advocated by the United Nations, "Make cities and human settlements inclusive, safe, resilient and sustainable." [Explanation of symbols]

[0108] 10. Camera 20 Management device (abnormality determination device) 100 Abnormality Judgment System 211 Interference determination section 213 Malfunction Judgment Section

Claims

1. An abnormality determination device that determines whether or not an abnormality occurs in an in-car camera that captures images inside a railway vehicle, An abnormality determination device comprising an interference determination unit that determines that imaging by the in-vehicle camera is being interfered with when a first condition is satisfied that the gain level value is within a normal range and the number of edges is below a predetermined first set value, using (i) a number of edges extracted by performing image processing on an image captured by the in-vehicle camera and (ii) a gain level value that is a numerical value of a gain amount by which an electrical signal generated in an imaging element of the in-vehicle camera when capturing the image, is amplified to a predetermined level.

2. a predetermined second set value higher than a normal range of the gain level value is set; 2. The abnormality determination device according to claim 1, wherein the interference determination unit determines that imaging by the in-vehicle camera is being interfered with when a second condition is satisfied that the gain level value is greater than the second set value, regardless of whether the first condition is satisfied.

3. The abnormality determination device according to claim 1 , wherein the obstruction determination unit determines that imaging by the in-vehicle camera is being obstructed when the state in which the first condition is satisfied continues for a predetermined time.

4. The abnormality determination device according to claim 2 , wherein the obstruction determination unit determines that imaging by the in-vehicle camera is being obstructed when the state in which the second condition is satisfied continues for a predetermined time.

5. 2. The abnormality determination device according to claim 1, wherein when all of the multiple gain level values ​​obtained at the same time from each of the multiple in-car cameras capturing images of the interior of the same railway vehicle exceed a predetermined third set value, the interference determination unit does not determine whether imaging by the in-car cameras is being interfered with.

6. 2. The abnormality determination device according to claim 1, further comprising a malfunction determination unit that, before a determination is made by the interference determination unit, compares the gain level values ​​obtained at the same time from each of the in-vehicle cameras capturing an image of the interior of the same railway vehicle, and determines that a malfunction has occurred in an in-vehicle camera having a gain level value whose difference from a median of the gain level values ​​is greater than a predetermined fourth set value.

7. The abnormality determination device according to claim 1 , wherein the number of edges is a number of vertical edges extracted by performing image processing on the captured image.

8. 8. An abnormality determination system comprising: the abnormality determination device according to claim 1; and the in-car camera installed in the railway vehicle.

9. An abnormality determination method for determining whether or not an abnormality has occurred in an in-car camera that captures images inside a railway vehicle, comprising: The abnormality determination method includes an interference determination step of determining that imaging by the in-vehicle camera is being interfered with when a first condition is satisfied that the gain level value is within a normal range and the number of edges is below a predetermined first set value, using (i) a number of edges extracted by performing image processing on an image captured by the in-vehicle camera and (ii) a gain level value that is a numerical value of a gain amount by which an electrical signal generated in an imaging element of the in-vehicle camera when capturing the image is amplified to a predetermined level.

Images