WO2021059339A1

WO2021059339A1 - Monitoring camera system

Info

Publication number: WO2021059339A1
Application number: PCT/JP2019/037335
Authority: WO
Inventors: 伊藤　哲也
Original assignee: 株式会社日立国際電気
Priority date: 2019-09-24
Filing date: 2019-09-24
Publication date: 2021-04-01
Also published as: JP7284280B2; JPWO2021059339A1

Abstract

Provided is a monitoring camera system with which it is possible to suppress any decrease in the visibility of a camera video in which the periphery of a car door is captured.　The monitoring camera system of the present example comprises: a camera 13 for imaging the periphery of the door of a train car 10, the camera 13 being attached to the outside of the car 10; a monitor 14 for displaying a camera video captured by the camera 13; and a control device 16 for assessing the visibility of the camera video on the basis of the feature quantity of the camera video and, when it is assessed that the visibility of the camera video has decreased, commanding the camera 13 to implement image processing (e.g., a luminance lowering process, a luminance heightening process, or a blur correction process) for improving the visibility of the camera video.

Description

Surveillance camera system

The present invention relates to a surveillance camera system that monitors the vicinity of a train door with a camera image.

Ensuring the safety of passengers when getting on and off the train is an important issue for train operation. As part of such safety measures, a vehicle monitoring CCTV (Closed Circuit Television) system has been developed that photographs the vicinity of the door with a camera attached to the side of the vehicle and displays the camera image on the monitor of the driver's cab (Closed Circuit Television). For example, see Patent Document 1).

By using such a system, the train driver can check the boarding and alighting of passengers on the monitor of the driver's cab when the train is stopped, and can operate the train. In the past, since real-time performance was required to monitor passenger safety, analog systems that did not cause delay were used with priority, but in recent years, advances in IP technology have made low-delay transmission possible. , All devices are now connected to the IP network.

JP-A-2018-113602

If the camera is mounted on the side of the vehicle, normal camera images may not be obtained due to backlight conditions such as sunlight, dust, dust, dirt, and mischief by passengers (for example, sticking foreign matter on the lens window). is there. As a result, the visibility of the camera image is lowered, and it may be difficult for the train driver to confirm the passengers getting on and off. The above-mentioned image defects are likely to occur when the camera is mounted on the outside of the vehicle, but there is a similar concern when the camera is mounted on the inside of the vehicle.

The present invention has been made in view of the above-mentioned conventional circumstances, and provides a surveillance camera system capable of suppressing a decrease in visibility of a camera image captured in the vicinity of a vehicle door. The purpose.

In the present invention, in order to achieve the above object, the surveillance camera system is configured as follows.
That is, the surveillance camera system according to the present invention is attached to the outside of the vehicle of the train and captures the vicinity of the door of the vehicle, a monitor that displays the camera image captured by the camera, and features of the camera image. Control that determines the visibility of the camera image based on the amount, and when it is determined that the visibility of the camera image has deteriorated, instructs the camera to execute image processing for improving the visibility of the camera image. It is characterized by being equipped with a device.

Here, the camera may be configured to analyze its own camera image, calculate a feature amount, and transmit the feature amount in association with the camera image.

Further, the control device may be configured to wait for an instruction to execute the image processing until the number of times it is determined that the visibility of the camera image has deteriorated exceeds a predetermined threshold value.

Further, the control device may be configured to record information on the determination result when it is determined that the visibility of the camera image has deteriorated.

Further, the control device may be configured to display the necessity of maintenance for the camera on the monitor based on the information of the determination result.

According to the present invention, it is possible to provide a surveillance camera system capable of suppressing a decrease in visibility of a camera image captured in the vicinity of a vehicle door.

It is a figure which shows the schematic configuration example of the surveillance camera system which concerns on one Embodiment of this invention. It is a figure which illustrates the flowchart of the image processing at the time of stopping at a station by the surveillance camera system of FIG. It is a figure which illustrates the flowchart of the feature amount processing in the image processing of FIG. It is a figure which illustrates the flowchart of the high-luminance correspondence processing in the feature amount processing of FIG. It is a figure which illustrates the flowchart of the low-luminance correspondence processing in the feature amount processing of FIG. It is a figure which illustrates the flowchart of the haze correspondence processing in the feature quantity processing of FIG. It is a figure which illustrates the histogram of a normal image. It is a figure which illustrates the histogram of the image of the haze state. It is a figure which illustrates the histogram of the image in a high brightness state. It is a figure which illustrates the histogram of the image in a low brightness state. It is a figure which illustrates the determination table used for the determination of the camera state by the surveillance camera system of FIG. It is a figure which shows the display example of the determination result of the camera state by the surveillance camera system of FIG.

A surveillance camera system according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 shows a schematic configuration example of an in-vehicle vehicle monitoring CCTV system, which is a surveillance camera system according to an embodiment of the present invention. FIG. 1 shows a 4-car train consisting of four cars, No. 1 car 10-1, No. 2 car 10-2, No. 3 car 10-3, and No. 4 car 10-4.

One switching hub 11 is arranged in each vehicle 10-1 to 10-4, and an IP network is constructed by cascading these. Further, on the outer side surface of each of the vehicles 10-1 to 10-4, a camera 13 is attached toward the door 12 of the vehicle so that a passenger can take an image of getting on and off the train. That is, the camera 13 is configured to take a picture of the vicinity of the door.

The leading vehicles, car 1 10-1 and car 4 10-4, are equipped with monitors 14 that display camera images taken by the cameras 13 of each vehicle. In addition, a control device 16 for controlling this system (vehicle monitoring CCTV) and a recording device 15 for recording camera images are also installed in the first car 10-1.

The control device 16 includes hardware resources such as a memory and a processor, and reads a program stored in a data storage device such as a hard disk or a flash memory onto the main memory and executes the program by the processor. It is configured to implement processing. The function of the control device 16 is not limited to the configuration realized by such software, and may be realized by dedicated hardware.

FIG. 2 illustrates a flowchart of video processing when the station is stopped in the surveillance camera system of this example. When the train stops at the station, the control device 16 performs the following processing. It is assumed that "0" is set as an initial value in the car number counter indicating the vehicle for which the image is to be acquired.

The control device 16 first determines whether or not the train door 12 is open (step S11), and when it is confirmed that the train door 12 is open, the car number counter is incremented (“1” is added) (step S12). ). Next, it is determined whether or not the car No. counter indicates the No. 5 car (step S13). Since car 5 does not exist, if the car counter indicates car 5 (step S13; Yes), the car counter is cleared (“1” is set in this example) (step S14). That is, the vehicle for which the image is to be acquired is looped in the order of the first car to the fourth car.

Next, the camera image is acquired from the camera 13 of the vehicle corresponding to the car number counter (steps S15 to S21). That is, when the car No. counter indicates the car No. 1 (step S15; Yes), the camera image is acquired from the camera 13 of the car No. 1 10-1 (step S18), and the car No. counter indicates the car No. 2 (step). S16; Yes) acquires the camera image from the camera 13 of the second car 10-2 (step S19), and when the car counter indicates the third car (step S17; Yes), it is from the camera 13 of the third car 10-3. The camera image is acquired (step S20), and in other cases (step S17; No), the camera image is acquired from the camera 13 of car No. 4 10-4 (step S21).

After acquiring the camera image, the control device 16 executes the feature amount processing (step S30) described later. After that, it is determined whether or not the train door 12 is closed (step S23), and if it is not closed (step S23; No), the process returns to the increment of the car number counter (step S12). On the other hand, when it can be confirmed that the train door 12 is closed (step S23; Yes), the process ends.

The feature amount processing (step S30) in the video processing of FIG. 2 will be described with reference to the flowchart of FIG.
In the feature amount processing, first, the control device 16 acquires the feature amount of the camera image from the header information of the camera image. In this example, the average luminance value and the luminance gradation width, which are a kind of feature quantities related to the luminance of the entire camera image, are acquired (steps S31 and S32).

The average brightness value is the average value of the brightness of each pixel constituting the camera image. The luminance gradation width is a luminance gradation width (range) in which the appearance frequency is equal to or higher than the reference value in the histogram of the appearance frequency of the brightness of each pixel constituting the camera image. These features are calculated by the camera 13 analyzing its own camera image, and are stored in the header information of the camera image. The feature amount is calculated by analyzing the camera image before performing image processing for improving visibility (low-luminance processing, high-luminance processing, haze correction processing, etc., which will be described later).

Next, the control device 16 executes high-luminance correspondence processing (step S40), low-luminance correspondence processing (step S60), and haze correspondence processing (step S80), respectively, based on the acquired feature amount of the camera image. The order in which these processes (steps S40, S60, S80) are executed is arbitrary.

The high-luminance processing (step S40) in the feature amount processing of FIG. 3 will be described with reference to the flowchart of FIG.
In the high-luminance processing, first, the control device 16 determines whether the low-luminance function of the camera 13 is ON or OFF (step S41). The low-luminance function is a function that executes low-luminance processing that lowers the brightness value of the entire image (all pixels) by image processing, and has been commercialized by various methods. Normally, since the low-luminance function is OFF, the processing in that case will be described first, and then the processing when the low-luminance function is ON will be described.

When the low brightness function is not ON (step S41; No), the control device 16 determines whether the condition that the average brightness value is larger than A1 and smaller than A2 (A1 <average brightness value <A2) is satisfied ( Step S42). As shown in the histogram example of FIG. 9, A1 and A2 are threshold values for determining whether the overall brightness of the image is abnormally high, and are set in advance. When this condition is satisfied, the overall brightness of the image is abnormally high, indicating that the image is overexposed.

When A1 <average brightness value <A2 (step S42; Yes), that is, when the camera image is in a high brightness state, the control device 16 executes the following processing.
First, the high-luminance counter is incremented (“1” is added) (step S43). The high-luminance counter counts the continuation status of the high-luminance state, and it is assumed that "0" is set as an initial value.

Next, it is determined whether or not the condition that the value of the high-luminance counter is larger than D1 (counter value> D1) is satisfied (step S44). D1 is a threshold value for providing a time grace so as not to repeat ON (start) and OFF (stop) of the low brightness function in a short time, and is set in advance. If the value of the high-brightness counter is larger than D1, after performing the high-brightness warning display process (step S45), the high-brightness recording process (step S46), and the low-brightness ON process (step S47), the high-brightness counter is displayed as ". 0 ”is set and initialized (step S48).

In the high-luminance warning display process (step S45), the control device 16 causes the monitor 13 to display that the average brightness of the image is high-luminance. In the high-luminance recording process (step S46), the control device 16 uses information indicating that the average brightness of the video is high-luminance as an event, and stores the event occurrence time and the occurrence camera ID (event generation source video) in its own storage area. Is recorded together with the generated camera identification information). In the low-brightness ON process (step S47), the control device 16 transmits a control signal for turning on the low-brightness function to the camera 13 that generated the event source image. As a result, the camera 13 in which the camera image is in the high brightness state starts the low brightness processing for the camera image.

When the low brightness function is ON (step S41; Yes), the control device 16 determines whether the condition that the average brightness value is larger than A1 and smaller than A2 (A1 <average brightness value <A2) is satisfied ( Step S52). In this example, the same thresholds as the thresholds A1 and A2 used in step S42 are used in step S52, but different thresholds A1'and A2' may be used.

When A1 <average luminance value <A2 (step S52; No), that is, when the camera image is in a non-luminance state, the control device 16 executes the following processing.
First, the non-brightness counter is incremented (“1” is added) (step S53). The non-brightness counter counts the continuation status of the non-brightness state, and it is assumed that "0" is set as an initial value.

Next, it is determined whether or not the condition that the value of the non-brightness counter is larger than D2 (counter value> D2) is satisfied (step S54). D2 is a threshold value for providing a time grace so that the low-luminance function is not repeatedly turned on and off in a short time, and is set in advance. D2 may have the same value as D1 or may have a different value from D1. If the value of the non-brightness counter is larger than D2, after performing the non-brightness recording process (step S56) and the low-brightness OFF process (step S57), the non-brightness counter is initially set to "0". (Step S58).

In the non-high-luminance recording process (step S56), the control device 16 records information indicating that the average brightness of the image is no longer high-luminance as an event in its own storage area together with the event occurrence time and the occurrence camera ID. .. In the low-brightness OFF process (step S57), the control device 16 transmits a control signal for turning off the low-brightness function to the camera 13 that generated the event source video. As a result, in the camera 13 where the camera image is no longer in the high brightness state, the low brightness processing for the camera image is stopped.

The low-luminance processing (step S60) in the feature amount processing of FIG. 3 will be described with reference to the flowchart of FIG.
In the low-luminance processing, first, the control device 16 determines whether the high-luminance function of the camera 13 is ON or OFF (step S61). The high-luminance function is a function that executes high-luminance processing for increasing the brightness value of the entire image (all pixels) by image processing, and has been commercialized by various methods. Normally, since the high-luminance function is OFF, the processing in that case will be described first, and then the processing in the case where the high-luminance function is ON will be described.

When the high brightness function is not ON (step S61; No), the control device 16 determines whether the condition that the average brightness value is larger than B1 and smaller than B2 (B1 <average brightness value <B2) is satisfied ( Step S62). As shown in the histogram example of FIG. 10, B1 and B2 are threshold values for determining whether the overall brightness of the image is abnormally low, and are set in advance. When this condition is satisfied, the overall brightness of the image is abnormally low, indicating that the image is dark.

When B1 <average brightness value <B2 (step S62; Yes), that is, when the camera image is in a low brightness state, the control device 16 executes the following processing.
First, the low-luminance counter is incremented (“1” is added) (step S63). The low-luminance counter counts the continuation status of the low-luminance state, and it is assumed that "0" is set as an initial value.

Next, it is determined whether or not the condition that the value of the low-luminance counter is larger than E1 (counter value> E1) is satisfied (step S64). E1 is a threshold value for providing a time grace so as not to repeat ON (start) and OFF (stop) of the brightness enhancement function in a short time, and is set in advance. If the value of the low-brightness counter is larger than E1, after performing the low-brightness warning display process (step S65), the low-brightness recording process (step S66), and the high-brightness ON process (step S67), the low-brightness counter is set to ". 0 ”is set and initialized (step S68).

In the low-luminance warning display process (step S65), the control device 16 causes the monitor 13 to display that the average brightness of the image is low. In the low-luminance recording process (step S66), the control device 16 uses information indicating that the average brightness of the video is low-luminance as an event, and stores the event occurrence time and the occurrence camera ID (event generation source video) in its own storage area. Is recorded together with the generated camera identification information). In the high-luminance ON process (step S67), the control device 16 transmits a control signal for turning on the high-luminance function to the camera 13 that generated the event source image. As a result, the camera 13 in which the camera image is in the low-luminance state starts the high-luminance process for the camera image.

When the high brightness function is ON (step S61; Yes), the control device 16 determines whether the condition that the average brightness value is larger than B1 and smaller than B2 (B1 <average brightness value <B2) is satisfied (B1). Step S72). In this example, the same thresholds as the thresholds B1 and B2 used in step S62 are used in step S72, but different thresholds B1'and B2' may be used.

When B1 <average luminance value <B2 (step S72; No), that is, when the camera image is in a non-luminance state, the control device 16 executes the following processing.
First, the non-low brightness counter is incremented (“1” is added) (step S73). The non-low-luminance counter counts the continuation status of the non-low-luminance state, and it is assumed that "0" is set as an initial value.

Next, it is determined whether the condition that the value of the non-low brightness counter is larger than E2 (counter value> E2) is satisfied (step S74). E2 is a threshold value for providing a time grace so as not to repeat ON and OFF of the brightness enhancement function in a short time, and is set in advance. E2 may have the same value as E1 or may have a different value from E1. If the value of the non-low brightness counter is larger than E2, after performing the non-low brightness recording process (step S76) and the high brightness OFF process (step S77), the non-low brightness counter is set to "0" for initialization. (Step S78).

In the non-low-luminance recording process (step S76), the control device 16 records information indicating that the average brightness of the image is no longer low-luminance as an event in its own storage area together with the event occurrence time and the occurrence camera ID. .. In the high-luminance OFF process (step S77), the control device 16 transmits a control signal for turning off the high-luminance function to the camera 13 that generated the event source image. As a result, in the camera 13 in which the camera image is no longer in the low brightness state, the high brightness processing for the camera image is stopped.

The haze-corresponding process (step S80) in the feature amount process of FIG. 3 will be described with reference to the flowchart of FIG.
In the haze handling process, first, the control device 16 determines whether the haze correction function of the camera 13 is ON or OFF (step S81). The haze correction function is a function that executes a haze correction process for correcting gradation by image processing so that a histogram as shown in FIG. 8 becomes a histogram as shown in FIG. 7, and has been commercialized by various methods. .. Normally, the haze correction function is OFF, so the processing in that case will be described first, and then the processing when the haze correction function is ON will be described.

When the haze correction function is not ON (step S81; No), the control device 16 determines whether or not the condition that the luminance gradation width is smaller than C (luminance gradation width <C) is satisfied (step S82). The luminance gradation width is calculated by, for example, subtracting the minimum value L1 from the maximum value L2 of the luminance whose appearance frequency exceeds H1 (see the histogram examples of FIGS. 7 and 8). C is a threshold value for determining whether or not the image is in a haze state (hereinafter referred to as “haze state”), and is set in advance. The haze state occurs not only when it is covered with haze, but also when the lens window of the camera is dirty or when it is backlit by sunlight. When this condition is satisfied, that is, when the luminance gradation width is narrow as shown in the histogram example of FIG. 8, it indicates that the image is in a haze state.

When the luminance gradation width <C (step S82; Yes), that is, when the camera image is in a haze state, the control device 16 executes the following processing.
First, the haze state counter is incremented (“1” is added) (step S83). The haze state counter counts the continuation status of the haze state, and it is assumed that "0" is set as an initial value.

Next, it is determined whether or not the condition that the value of the haze state counter is larger than F1 (counter value> F1) is satisfied (step S84). F1 is a threshold value for providing a time grace so as not to repeat ON (start) and OFF (stop) of the haze correction function in a short time, and is set in advance. If the value of the haze state counter is larger than F1, after performing the haze state recording process (step S86) and the haze correction ON process (step S87), the haze state counter is set to "0" and initialized (step). S88).

In the haze state recording process (step S86), the control device 16 uses the information indicating that the image is in the haze state as an event, and generates the event occurrence time and the occurrence camera ID (event generation source image) in its own storage area. Record with camera identification information). In the haze correction ON process (step S87), the control device 16 transmits a control signal for turning on the haze correction function to the camera 13 that generated the event source image. As a result, the camera 13 in which the camera image is in the haze state starts the haze correction process for the camera image.

When the haze correction function is ON (step S81; Yes), the control device 16 determines whether or not the condition that the luminance gradation width is smaller than C (luminance gradation width <C) is satisfied (step S92). In this example, the same threshold value as the threshold value C used in step S82 is used in step S92, but another threshold value C'may be used.

When the luminance gradation width <C (step S92; No), that is, when the camera image is in a non-haze state, the control device 16 executes the following processing.
First, the non-haze state counter is incremented (“1” is added) (step S93). The non-haze state counter counts the continuation status of the non-haze state, and it is assumed that "0" is set as an initial value.

Next, it is determined whether or not the condition that the value of the non-haze state counter is larger than F2 (counter value> F2) is satisfied (step S94). F2 is a threshold value for providing a time grace so that the haze correction function is not repeatedly turned on and off in a short time, and is set in advance. F2 may have the same value as F1 or may have a different value from F1. If the value of the non-haze state counter is larger than F2, after performing the haze correction OFF recording process (step S56) and the haze correction OFF process (step S97), set “0” to the non-haze state counter and initialize it. (Step S98).

In the haze correction OFF recording process (step S96), the control device 16 records information indicating that the image is not in the haze state as an event in its own storage area together with the event occurrence time and the occurrence camera ID. In the haze correction OFF process (step S97), the control device 16 transmits a control signal for turning off the haze correction function to the camera 13 that generated the event source image. As a result, the haze correction process for the camera image is stopped at the camera 13 in which the camera image is no longer in the haze state.

As described above, when the train station is stopped, the control device 16 determines the image defect (decreased visibility) based on the feature amount of the image of the camera 13, and improves the visibility according to the detected image defect. A control signal instructing the execution of the process is transmitted to the camera 13.

After that, when the train operation is completed and the train is maintained, the control device 16 records by the above-mentioned high-luminance / non-high-luminance recording, low-luminance / non-low-luminance recording, haze correction ON / OFF recording, or the like. Based on the received information, the state of the camera 13 is determined, and the determination result is displayed on the monitor 14 to notify the maintenance staff. The control device 16 may display the information itself recorded by the high-luminance / non-high-luminance recording, the low-luminance / non-low-luminance recording, the haze correction ON / OFF recording, and the like on the monitor 14.

FIG. 11 illustrates a determination table used for determining the camera state. In the judgment table of FIG. 11, when the haze correction function is ON for a long time, the camera state is determined as "dirty", and when the high brightness function is ON for a long time, the camera state is defined as "lens window blockage". It is stipulated that the camera state is determined as "camera abnormality" when the low brightness function is ON for a long time, and the camera state is determined as "normal" when none of the above applies. That is, the control device 16 of this example calculates the ON state time for each of the haze correction function, the high brightness function, and the low brightness function, and depending on whether or not the ON state time is equal to or greater than a predetermined threshold value. The state of the camera 13 is determined.

FIG. 12 shows a display example of the determination result of the camera state. As shown in the figure, the monitor 14 displays information indicating the camera state (that is, information indicating the need for maintenance) for each of the cameras 13 mounted on the train. Therefore, by checking the display on the monitor 14 when servicing the train, the maintenance staff can easily determine what kind of camera state each of the cameras 13 mounted on the train is in and the need for maintenance. It becomes possible to recognize.

As described above, the surveillance camera system of this example includes a camera 13 which is attached to the outside of the vehicle 10 of the train and images the vicinity of the door of the vehicle 10, a monitor 14 which displays a camera image captured by the camera 13. The visibility of the camera image is determined based on the feature amount of the camera image, and when it is determined that the visibility of the camera image is deteriorated, image processing (low brightness processing, high brightness) for improving the visibility of the camera image is performed. It is provided with a control device 16 that instructs the camera 13 to execute the conversion process, the haze correction process, and the like.

Therefore, according to the surveillance camera system of this example, the camera 13 automatically executes image processing corresponding to the mode of the decrease in visibility according to the decrease in the visibility of the camera image captured in the vicinity of the door of the vehicle. .. As a result, it becomes possible to prevent the visibility of the camera image from being lowered and the train driver from having trouble in confirming the boarding and alighting of passengers.

Further, in the surveillance camera system of this example, the camera 13 has a configuration in which the camera 13 analyzes its own camera image, calculates a feature amount, and transmits the feature amount along with the camera image. Therefore, since the feature amount can be calculated from the camera image (original camera image) before the image processing for improving the visibility of the camera image is performed, there is no concern that the feature amount of the camera image will be changed by the image processing. Further, the processing load of the control device 16 can be reduced.

Further, in the surveillance camera system of this example, the control device 16 gives an instruction to execute image processing for improving the visibility of the camera image until the number of times it is determined that the visibility of the camera image has deteriorated exceeds a predetermined threshold value. It is configured to wait for. Therefore, it is possible to prevent the start and end of the image processing from being repeated in a short time.

Further, in the surveillance camera system of this example, when it is determined that the visibility of the camera image is deteriorated, the control device 16 is configured to record the information of the determination result. Therefore, for example, the maintenance staff can confirm the state of deterioration of the visibility of the camera image after the fact (after the train is stopped, etc.).

In the surveillance camera system of this example, the control device 16 is configured to display the necessity of maintenance for the camera 13 on the monitor 14 based on the information of the determination result. Therefore, for example, the maintenance staff can easily recognize the necessity of maintenance for the camera 13 such as cleaning the lens window only by checking the display on the monitor 14.

In the above explanation, a 4-car train was used as an example, but it goes without saying that the number of trains is arbitrary. Further, in the above description, the control device 16 has a function of controlling the camera 13 based on the feature amount of the camera image, but another device (for example, the monitor 14 or the recording device 15) has the function. It may be configured. Further, in the above description, the camera 13 is configured to perform image processing for improving the visibility of the camera image, but the image processing is performed by another device (for example, a monitor 14, a recording device 15 or a control device 16). May be the configuration performed by.

Although the present invention has been described in detail above, it goes without saying that the present invention is not limited to the above configuration and may be realized by a configuration other than the above. For example, the present invention can be applied to various types of trains such as monorails and trams, in addition to trains traveling on railways.

Further, the present invention provides, for example, a method or method for executing a process according to the present invention, a program for realizing such a method or method by a computer having hardware resources such as a processor or memory, and such a program. It can also be provided as a storage medium for storage.

The present invention can be used in a surveillance camera system that monitors the vicinity of a train door with a camera image.

10-1, 10-2, 10-3, 10-4: Vehicle, 11: Switching hub, 12: Door, 13: Camera, 14: Monitor, 15: Recording device, 16: Control device

Claims

A camera mounted on the outside of the train car that captures the vicinity of the car door,
A monitor that displays the camera image captured by the camera, and
The visibility of the camera image is determined based on the feature amount of the camera image, and when it is determined that the visibility of the camera image is deteriorated, the execution of image processing for improving the visibility of the camera image is performed. An in-vehicle surveillance camera system characterized by being equipped with a control device that instructs the camera.
In the surveillance camera system according to claim 1,
The camera is a surveillance camera system characterized in that it analyzes its own camera image, calculates a feature amount, and transmits the feature amount in association with the camera image.
In the surveillance camera system according to claim 1 or 2.
The control device is a surveillance camera system characterized in that it waits for an instruction to execute the image processing until the number of times it is determined that the visibility of the camera image has deteriorated exceeds a predetermined threshold value.
In the surveillance camera system according to any one of claims 1 to 3.
The control device is a surveillance camera system characterized in that when it is determined that the visibility of the camera image is deteriorated, information on the determination result is recorded.
In the surveillance camera system according to claim 4.
The control device is a surveillance camera system characterized in that the monitor displays the necessity of maintenance for the camera based on the information of the determination result.