JP2011063416A - Violence detector and violence detecting method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a violence detector which detects violence of a person in an elevator car, without erroneous detection, and detects precisely the violence of the person. <P>SOLUTION: This violence detector includes a reception part for receiving an image inside the car photographed by an imaging device, a storage part for storing the received images classifiedly by dates and times, a dividing means for dividing the stored image 26 and the received image 27 respectively into a plurality of small areas, an analytical part for weighting the small area with a change generated between the consecutive images, out of the stored image 26 having the plurality of divided small areas, and for weighting the small area with a change generated between the consecutive images, out of the received image 27 having the plurality of small areas, and for generating a digitized data obtained by digitizing each image, and a detecting part for calculating the differential amount between the two digitized data generated by the analytical part, and for detecting the small area having the differential amount larger than a threshold value, as an area with the violence of the person, out of the images inside the car from the reception part, based on the differential amount and the prestored threshold value. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a ramp detection device and a ramp detection method provided in an elevator car or the like.

  In recent years, elevators installed in condominiums and the like have been increasingly equipped with a function of recording a video by installing a security camera in a car for the purpose of crime prevention. The rampage detection device is a device that uses a video from a security camera in a cage to detect that a person is rampant in the cage from the amount of change in the motion in the video and the rate of change of this motion. A ramp detection device with improved security is known by combining the ramp detection with the ramp detection device with the announcement in the car room or the control of stopping the car to the nearest floor (see Patent Document 1).

JP 2006-276969 A

  However, in the conventional rampage detection device, the rampage detection method uses the amount of change in the motion of the camera image and the rate of change in motion, and therefore erroneous detection may occur due to a change in the image other than the rampage that should be detected originally. is there. For example, if the elevator equipment is equipped with a security window on the car door, or the hoistway wall is glazed, the change in the sun, the change in the external environment that can be seen from the security window when the elevator car goes up or down, or the brightness in the car Due to this change, there is a possibility that a state where a person is rampant or a person stays in error even though no one is actually in the car room.

  In view of the above problems, an object of the present invention is to provide a ramp detection device that can avoid erroneous detection of the ramp detection device and can detect a rampage of a person with high accuracy.

  In order to solve such a problem, according to one aspect of the present invention, a receiving unit that receives an image in the car imaged by an imaging device that images the elevator car, and an image received by the receiving unit. A storage unit that stores data by date and time, a division unit that divides the video stored in the storage unit and the video received by the reception unit into a plurality of small regions, and a plurality of divisions divided by the division unit In the video of the storage unit having a small area, weighting is applied to the small area in which a change in motion occurs between successive images, and movement is performed between successive images in the video of the receiving unit having a plurality of small areas. A weight is applied to the small area where the change occurs, and an analysis unit that generates digitized data obtained by digitizing each video and a difference amount between the two digitized data generated by the analysis unit are calculated. Memorize the amount in advance A detection unit that detects, based on a threshold value, a small area having a difference amount larger than the threshold value from a video image in the car from the reception unit as an area where a person is rampant. A rampage detection device is provided.

  According to another aspect of the present invention, a storage unit that stores a plurality of images obtained by imaging an image in an elevator car by date and time by an imaging device, and an interior of the elevator car Means for obtaining a current image by imaging with the imaging device, means for extracting a partial region that changes with time from among the partial regions in the current image, and date and time for the current image Means for selecting a corresponding video in the storage unit as a reference video, means for extracting a partial area that changes with time from among the partial areas in the selected reference video, and a partial area extracted from the current video And a means for comparing the partial area extracted from the reference area, and a means for outputting a partial area that changes with time in the current video, the output means comprising a comparison result. , Rage detecting device is provided, characterized in that to suppress the output of the partial region indicating the same change.

  According to another aspect of the present invention, the computer stores (a) a plurality of videos obtained by imaging the images in the elevator car by date and time by the imaging device in the storage unit; (B) The imaging device images the inside of the elevator car and stores the current video in the storage unit, and (c) extracts a partial area that changes with time from among the partial areas in the current video. , (D) selecting a video in the storage unit corresponding to the date and time as the reference video for the current video, and (e) a partial area that changes with time among the partial areas in the selected reference video (F) comparing the partial area extracted from the current video with the partial area extracted from the reference area, and (g) outputting a partial area that changes with time in the current video. Configured (H) a result of the comparison, rage detecting method characterized by inhibiting the output of the partial region indicating the same change is provided.

  According to the present invention, it is possible to detect a human rampage in an elevator car without erroneous detection, and to detect a human rampage with high accuracy.

It is a lineblock diagram of an elevator which has a rampage detection device concerning a 1st embodiment of the present invention. 1 is a block diagram of a ramp detection device according to a first embodiment of the present invention. (A) is a figure for demonstrating the processing method which calculates | requires the variation | change_quantity for every small area about the past image, (b) is for demonstrating the processing method which calculates | requires the variation | change_quantity for every small area about the present image | video. FIG. It is a figure for demonstrating the method of the comparison between videos. It is a block diagram of the rampage detection apparatus which concerns on the 2nd Embodiment of this invention. It is a block diagram of the rampage detection apparatus which concerns on the 4th Embodiment of this invention.

  Hereinafter, a ramp detection device according to an embodiment of the present invention will be described with reference to FIGS. 1 to 6. In the drawings, the same portions are denoted by the same reference numerals, and redundant description is omitted.

(First embodiment)
FIG. 1 is a configuration diagram of an elevator having a ramp detection device according to a first embodiment of the present invention. The elevator 1 includes a car 3 installed in the hoistway 2, a hoisting machine 5 around which the main rope 4 is wound, a counterweight 6 provided in the hoistway 2, and an elevator that controls the operation of the car 3. And a control device 7.

  In the car 3, an in-car operation panel 8 is provided on the wall surface beside the doorway, and a security camera device 9 is attached to the ceiling surface of the car room. A car control device 10 that is electrically connected to the in-car operation panel 8 and the camera device 9 is provided on the roof of the car 3, and the car detection device 11 according to the present embodiment is provided in the car control device 10. Has been implemented. This rampage detection device 11 is a device that detects that the passenger is rampant in the car from the amount of change in motion in the video and the rate of change of this motion. The rampage detection device 11 notifies the car control device 10 of a detection signal when the rampage of the passenger is detected.

  The in-car operation panel 8 is provided with a display 12 and an announcement device 13 such as a speaker. When the car control device 10 receives the detection signal from the ramp detection device 11, the car control device 10 notifies the display device 12 and the announcement device 13 of an alarm. The car control device 10 is also connected to the elevator control device 7 by a tail cord 14, and a control signal can be transmitted between the car control device 10 and the elevator control device 7. When the elevator control device 7 receives a control signal indicating that the rampage has been detected from the car control device 10, the elevator control device 7 causes the hoist 5 to run the car 3 to the nearest floor and stop it. ing.

  The car 3 has a car door 15, and the car door 15 is provided with a security window 16 such as a see-through glass. The elevator hall 17 on each floor is provided with an entrance / exit and a hold door 18 that opens and closes each entrance / exit. Each hold door 18 is also provided with a security window 19. The hold door 18 is transmitted with door opening and closing driving force from the car door 15 by an engagement device (not shown) provided on the car door 15, and the door opening is performed in synchronization with the door opening or closing of the car door 15. Or close the door.

  Since the outside light and illumination in the elevator hall 17 enter the car room through the security windows 19 and 16, the brightness of the car room is changed by raising and lowering the car 3 and opening and closing the car door 16. The brightness in the captured image also changes. There is a case where an electrical decoration device for advertisement or as a signboard is installed indoors or outdoors in a building where the elevator 1 is installed. The electrical decoration device may blink and change color due to a highly decorative light emission pattern effect. The ramp detection device 11 according to the present embodiment is configured such that no erroneous detection operation occurs even when flashing or discolored light is inserted into the car 3 through the security windows 19 and 16. Hereinafter, the ramp detection device 11 will be described in detail.

  FIG. 2 is a block diagram of the ramp detection device 11 according to the present embodiment. Of the reference signs shown in the figure, the reference signs already represent the same elements. The rampage detection device 11 stores the calendar control unit 20 that outputs the current date information and time information, and the camera device 9 for one year, which is obtained by photographing the interior of the car 3 every 365 days a year. When the camera image storage unit 21 (storage unit) and the video obtained from the camera video storage unit 21 and the current video from the camera device 9 are image-analyzed, and an area where movement occurs between the videos is detected, It includes a camera video analysis unit 22 (reception unit, analysis unit, detection unit) that detects that a person is rampant. The camera device 9 mounted in the car 3 is connected to the camera image analysis unit 22 via the communication line 23. The camera video analysis unit 22 is connected to the camera video storage unit 21 via the communication line 24. The camera video analysis unit 22 is connected to the calendar control unit 20 via the communication line 25.

  As an example, the calendar control unit 20 includes a crystal oscillator (not shown), a voltage source that generates a voltage applied to the crystal oscillator, a counter circuit that counts the oscillation frequency of the crystal oscillator, and a counter circuit that counts the count frequency. A storage circuit that stores data such as time and a reading unit that reads information such as date and time from the storage circuit are provided.

  The camera video storage unit 21 is a storage device that stores video data for each date and time for one year. The camera video storage unit 21 uses a large-capacity hard disk drive provided on the car roof. The video data for each date and each time is a piece of video data that is dropped every few seconds.

  The camera video analysis unit 22 determines the whole or part of the screen area of the input video from the camera device 9 as the video area to be analyzed, divides this video area into a plurality of small areas, and reproduces two videos. The input video is digitized by weighting one or more small areas that have changed since then. The camera video analysis unit 22 performs the same analysis on the input video from the camera video storage unit 21 as the analysis on the input video from the camera device 9.

  In addition, the camera video analysis unit 22 compares the input video from the camera device 9 composed of a plurality of quantified small regions with the input video from the camera video storage unit 21 composed of a plurality of quantified small regions. Then, the difference between each small area of these input images is calculated, and if any of the difference values exceeds a predetermined threshold, the small area having this difference value is identified as the area where the motion has occurred. To do. By specifying the area, a local change in the video region to be analyzed is detected.

  The function of the camera image analysis unit 22 is executed by a computer having a CPU, a ROM, a RAM, and a communication interface. By cooperating with these CPUs and the like, functions of video reception, video analysis, and motion detection in the video are realized.

  The operation of the rampage detection apparatus 11 according to this embodiment configured as described above will be described. The camera device 9 installed in the car 3 always shoots the video in the car 3 and transmits the video taken to the camera video analysis unit 22 via the communication line 23.

  The camera video analysis unit 22 acquires the current date and time from the calendar control unit 20 via the communication line 25, and transmits the acquired date and time information to the camera video storage unit 21 via the communication line 24.

  The camera video storage unit 21 stores in advance a plurality of videos taken in the car 3 by date and time. The computer searches the camera video storage unit 21 from the date and time information received from the camera video analysis unit 22, selects the same date and time video from the stored video, and selects the same date. The time image is transmitted to the camera image analysis unit 22 via the communication line 24.

  The camera image analysis unit 22 performs analysis such as extraction of the change direction of the image obtained from the camera image storage unit 21 and measurement of the change speed, and weights which area in the image has changed. To get it. The change amount detection method will be described with reference to FIGS.

  FIG. 3A is a diagram for explaining a processing method for obtaining a change amount for each small area in the past video. The camera video analysis unit 22 reads the beginning of the video as a reference video at time t0 on the RAM. The camera image analysis unit 22 equally divides these images into, for example, 16 small regions by parallel vertical lines and parallel horizontal lines. The camera video analysis unit 22 quantifies how much the video at these times has changed from the reference video at each time such as t1, t2,. As a numerical method, a method for obtaining a change amount for each video area such as an “optical flow method” is used. The optical flow is a speed component of the motion on the video obtained by the arithmetic processing. The camera image analysis unit 22 obtains a vector by connecting points between the reference image at t0 and the other images at t1 to tn. The magnitude of this vector corresponds to the velocity component of movement. Or the camera image | video analysis part 22 calculates | requires a vector by connecting the partial area | region which consists of a some pixel between a reference | standard image | video and another image | video.

  FIG. 4 is a diagram for explaining a method of comparison between videos. In FIG. 4, the camera video analysis unit 22 compares the reference video 26 at time t0 with the video 27 at any one of times t1 to tn. The camera image analysis unit 22 generates a weight data table 28 having 16 data areas corresponding to the divided 16 small areas on the RAM. The camera video analysis unit 22 gives a weight to a data area corresponding to the changed area in the weight data table 28, for example, when the upper right of the video changes even once. When the weight value is represented by binary values 1 and 0, the camera image analysis unit 22 sets a bit representing the weight in the data area. In addition, the change width of the image may be divided into the weight value addition widths finely according to the change speed and change time of the brightness. As a method that can be used by the camera video analysis unit 22, a method other than the optical flow method may be used as long as the amount of change is obtained for each video area.

  Next, the camera video analysis unit 22 analyzes the current video sent from the camera device 9 using the same method, weights it, and digitizes it.

  FIG. 3B is a diagram for explaining a processing method for obtaining a change amount for each small area of the current video from the camera device 9. The camera video analysis unit 22 uses the beginning portion of the current video as a reference video, and quantifies the change of the current video at each time such as times t1, t2,. The camera image analysis unit 22 divides the current image into 16 parts. The presence / absence of a vector is determined by connecting points between points or a plurality of pixels between the reference image at t0 in FIG. 3B and the current images at t1 to tn.

  Subsequently, the camera video analysis unit 22 compares the numerical data obtained by analyzing and digitizing the current video with the data obtained by analyzing and digitizing the video stored in the camera video storage unit 21, After calculating the data difference, the difference data for the rampage determination is obtained. When the camera image analyzing unit 22 detects an area that changes more than a certain value from the difference data for the rampage determination, it outputs a detection signal as the rampage detection. When the elevator control device 10 receives this detection signal, the elevator control device 10 performs operations such as sounding the announcement device 13 in the car 3.

  In this way, the camera video analysis unit 22 makes a difference in the amount of change between the initial past reference video used for comparison with respect to the past image and the subsequent past video, and the initial current reference used for comparison with respect to the current image. The difference between the video and the current video after that is compared.

  Therefore, in this rampage detection device 11, even if there is a change in the sunlight from the outside, a change in the lighting of the store, etc., by comparing the difference in the past change amount with the difference in the current change amount, It is possible to suppress changes in the video. By the processing of the ramp detection device 11, changes in the image other than the ramp detection target are suppressed as much as possible, and more accurate detection is possible in the detection of an abnormal situation in the car. According to the ramp determination device 11 according to the present embodiment, it is possible to detect a rampage of a person in the car 3 of the elevator 1 without erroneous detection, and it is possible to detect a rampage of a person with high accuracy.

  Since the video for each date is stored in the camera video storage unit 21, the camera video analysis unit 22 can select a reference video for video comparison even when the length of daytime and the length of night change. Is possible. Since the video is stored for each date, the camera video storage unit 21 can output a reference video as a season-specific video such as a winter season from December to March and a spring season from March to June. . Detection processing can be performed in consideration of changes in brightness that change according to the length of the day, and false detection is reduced.

  In addition, for example, since video for each time such as one hour interval is stored, the camera video analysis unit 22 can prepare a reference video for comparison for each time zone. In condominiums, there are many users in the morning and there are many people. The video shot is also different from the video shot during the daytime. The business hours of the store and the time zone during which the signs are lit up vary depending on the time of day. The camera image analysis unit 22 can perform detailed comparison analysis.

  The degree of change at night is larger than the degree of change obtained by comparing images under bright illumination during the day. Since the video by time is stored, the camera video analysis unit 22 can compare using a large number of videos such as the morning reference video and the night reference video, and the length of the day and the length of the night can be compared. Even if it changes, analysis processing can be performed using an appropriate reference image. Therefore, it is possible to reduce false detections.

  There is an ornament with a colored bean bulb indoors as an illumination device. LED lamps and neon tubes are used for advertisements and billboards installed on the outer walls of buildings and shopping streets. For example, when the car 3 ascends and descends on a glass hoistway, the magnitude of changes in outside light and scenery by the electrical decoration device is large. According to the rampage detection apparatus 11 according to the present embodiment, false detection can be reduced. In the rampage detection system for detecting an abnormal situation in the car 3, more accurate detection can be realized, and appropriate notification can be performed.

(Second Embodiment)
In the second embodiment, the rampage detection device selects the reference video according to the weather.

  FIG. 5 is a block diagram of a ramp detection apparatus according to the second embodiment of the present invention. In the figure, the above-mentioned reference numerals represent the same elements as those described above.

  The configuration of the ramp detection device 29 according to the present embodiment is connected to the camera video storage unit 21 in the calendar control unit 20, the camera video analysis unit 22, and the camera video storage unit 21 in the ramp detection device 11 of FIG. In this configuration, a communication line 30 and a weather detection unit 31 connected to the communication line 30 are added.

  The weather detection unit 31 outputs current weather information such as sunny weather outside the building and cloudy weather. For the weather detection unit 31, a communication terminal connected to the Internet is used. Alternatively, a rain / snow sensor that senses rainfall or snow and outputs a sensing signal is used for the weather detection unit 31.

  The weather information detection method performed by the weather detection unit 31 of the ramp detection device 29 according to the present embodiment having the above-described configuration is obtained from a weather information distribution service by Internet distribution, a rain / snow sensor, or the like periodically. A method of receiving a signal sent to is used.

  The camera video storage unit 21 receives weather information sent from the weather detection unit 31 via the communication line 30, and stores a plurality of videos having the same date and time as the received weather information. For example, at 6:00 pm on March 31, the camera image storage unit 21 stores in advance the images in the car 3 in the case of clear weather, cloudy weather, snowfall, and rain. The camera video storage unit 21 transmits all stored videos to the camera video analysis unit 22 in response to a request from the camera video analysis unit 22.

  The camera video analysis unit 22 uses the same method as the video analysis method in the first embodiment. The camera image analysis unit 22 obtains a change amount for the current image sent from the camera device 9. The camera video storage unit 21 sends a plurality of reference videos selected according to a plurality of weather types to the camera video analysis unit 22. The camera video analysis unit 22 calculates the amount of change for all the videos sent from the camera video storage unit 21. The camera image analysis unit 22 compares the amount of change obtained for the current image with the amount of change obtained for the image sent in accordance with each weather type. And so on.

  Therefore, according to the rampage detection device 29 according to this embodiment of the present invention, the reference image is stored in advance as images with different weather at the same date and time, and is compared with all the images so that the difference in weather can be detected. Even if changes, it is possible to suppress erroneous detection of rampage.

  If there is a single reference video for the same date and time, there may be a case where a certain detection accuracy cannot be maintained in the process of comparing the current video with the reference video. The rampage detection device 29 according to the present embodiment can obtain a change amount for a past image in a clear sky, a change amount for a past image in a cloudy weather, and the like by preparing a plurality of reference images for the same date and time. become able to. Since the camera image analysis unit 22 compares the amount of change obtained from the current image with the amount of change for each weather, the certainty of detection can be achieved.

  In this embodiment, the weather is selected as the type of the reference video, but a plurality of types of reference video may be prepared using conditions other than the weather. When a person crosses in front of the security window 16 or the security window 19 at the timing when the camera device 9 is photographed, if the reference image is single, a false detection is caused by the person who has accidentally passed. In order to avoid this erroneous detection, the camera image analysis unit 22 can perform detection determination using a plurality of types of stored images, thereby improving detection accuracy.

(Third embodiment)
In the third embodiment, in the first embodiment and the second embodiment, when the detection time of the difference between videos detected by the ramp detection device 11 or the ramp detection device 29 continues for a certain time or longer, the reference video Is to be updated. The camera image analysis unit 22 of the rampage detection apparatus 11 in FIG. 2 has a time-lapse measurement function realized by a CPU timer.

  With such a configuration, the camera video analysis unit 22 starts the timer counter operation in advance. When the camera image analysis unit 22 compares the amount of change obtained for the past image from the camera image storage unit 21 with the amount of change obtained for the current image from the camera device 9, the camera image analysis unit 22 exceeds the set time set in the timer, When the same difference is detected, the video data sent from the camera device 9 is transmitted to the camera video storage unit 21. The camera video storage unit 21 stores the received video in a storage device in the camera video storage unit 21 and updates the reference video for comparison.

  Also, the camera image analysis unit 22 of the rampage detection device 29 of FIG. 4 performs the same operation as this example and has the same effect.

  Therefore, according to the rampage detection device according to this embodiment of the present invention, when a video that is not stored in the camera video storage unit 21 is continuously shot by the camera device 9, the rampage can be detected by updating the reference video. It becomes possible to suppress erroneous detection.

  When the elevator is first operated in the building and the ramp detection device 22 starts to turn on the power of the camera device 9, the ramp detection device 22 starts a number of shootings and collects reference images. For example, when a new building is constructed in the direction facing the elevator hall window after the elevator operation continues, the shadow of the building is inserted into the security window 16 only during a fixed time of the day. If this is left unattended, the rampage detection function always operates at a specific time every day. In the rampage detection apparatus according to this embodiment, erroneous detection is avoided by updating the reference video as necessary. It is possible to prevent the occurrence of a situation in which changes always occur for 1-2 hours in the morning every day.

(Fourth embodiment)
In 4th Embodiment, the false detection of the rampage detection apparatus which may be caused by the door opening operation | movement of a car door and a door closing operation | movement is avoided.

  FIG. 6 is a block diagram of a ramp detection apparatus according to the fourth embodiment of the present invention. In the figure, the above-mentioned reference numerals represent the same elements as those described above.

  The rampage detection device 32 includes a calendar control unit 20, a camera video analysis unit 22 </ b> A, and a camera video storage unit 21. The camera image analysis unit 22A is connected to the elevator control device 7 via the tail cord 14 and the car control device 10 (not shown). Other than this point, the camera video analysis unit 22A has the same function as the camera video analysis unit 22 has.

  The car control device 10 is configured to receive a signal for instructing to open the door from the elevator control unit device 7. Signals indicating the open / closed state of the car door are notified from the car control device 10 to the elevator control device 7 via the tail cord 14. The elevator controller 7 can monitor the state of the car 3.

  The ramp detection device 32 according to the present embodiment receives signals from the elevator control device 7 during the raising / lowering of the car 3 and the opening / closing operation of the car door, and invalidates the ramp detection function when the elevator equipment is operating. Processing to do. Thereby, the ramp detection device 32 suppresses erroneous detection.

  The elevator control device 7 that exchanges data with the rampage detection device 32 according to the present embodiment having such a configuration causes the elevator car 3 to travel and controls the opening and closing of the car door. During operation, the elevator control device 7 transmits a signal indicating that the rampage detection operation is prohibited to the camera image analysis unit 22A via the tail cord 14.

  The camera video analysis unit 22 performs processing so as not to detect the occurrence of a ramp even if any video change is detected while the ramp detection operation prohibition signal is received.

  Therefore, according to the violence detection device 32 according to this embodiment of the present invention, the image of the image generated by the operation of the elevator device can be changed against the change of the external environment during the elevator running reflected from the security window 16 and the change of the image when the car door is opened and closed. When a change occurs, the rampage detection function is disabled, so that it is possible to suppress false detection of rampage.

  In this ramp detection device 32, instead of directly performing detection processing using the current video from the camera device 9, this signal prohibits ramp detection operation processing using an elevator operation signal from the elevator control device 7 side. Since the detection function is cut only during the operation, software implementation can be performed relatively easily by utilizing the functions already used in actual elevator operation.

(Other embodiments)
Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage.

  In the above embodiment, the ramp detection device 11 is mounted on the circuit boards housed in the box provided with the roof of the car 3, but the configuration of the ramp detection device 11, for example, a reference image It is possible to make various changes such as making the storage device for storing a separate configuration via connectors. The superiority of the present invention is not lost to the invention and the implementation of the invention which have only been made such changes.

  In addition, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment.

  DESCRIPTION OF SYMBOLS 1 ... Elevator, 2 ... Hoistway, 3 ... Car, 4 ... Main rope, 5 ... Hoisting machine, 6 ... Counterweight, 7 ... Elevator control device, 8 ... In-car operation panel, 9 ... Camera device (imaging device) DESCRIPTION OF SYMBOLS 10 ... Car control apparatus, 11, 29, 32 ... Rampage detection apparatus, 12 ... Display, 13 ... Announcement apparatus, 14 ... Tail cord, 15 ... Car door, 16, 19 ... Security window, 17 ... Elevator hall, 18 ... Holder, 20 ... Calendar control unit, 21 ... Camera video storage unit (storage unit), 22, 22A ... Camera video analysis unit (reception unit, analysis unit, detection unit), 23, 24, 25, 30 ... communication line , 26: Reference video (stored video), 27: Camera video (received video), 28: Weight data table, 31: Weather detection unit.

Claims (6)

A receiving unit that receives an image of the car imaged by an imaging device that images the elevator car;
A storage unit for storing video received by the receiving unit by date and time;
A dividing unit that divides each of the video stored in the storage unit and the video received by the receiving unit into a plurality of small regions;
In the video of the storage unit having a plurality of small areas divided by the dividing unit, weighting is performed on the small areas where a change in motion occurs between successive images, and the receiving unit having the plurality of small areas An analysis unit that performs weighting on a small area in which a change in motion occurs between successive images in a video, and generates digitized data in which each video is digitized;
The difference between the two digitized data generated by the analysis unit is calculated, and based on the difference and a threshold value stored in advance, a difference larger than the threshold value is selected from the images in the cage from the reception unit. A rampage detection apparatus comprising: a detection unit that detects a small area having a volume as a human rampage area.   The storage unit stores a plurality of images corresponding to the same date and the same time, the analysis unit generates the digitized data for all of these images, and the detection unit includes all the images and the images. 2. The rampage detection apparatus according to claim 1, wherein when a video difference is detected by comparing all the video images received by the receiving unit with the current video image in the car, the rampage detection device is detected.   When the change in the difference between the video received from the receiving unit and the video stored in the storage unit continues continuously for a preset time or longer, the analysis unit stores the video in the car stored in the storage unit The rampage detection device according to claim 1, wherein: An input means for receiving a control signal from an elevator control device for controlling the operation of the elevator;
When the analysis unit receives a control signal from the input means during raising / lowering of the car or opening / closing operation of the car door, the analysis unit has a rampage detection function that the analysis unit has while the elevator apparatus is operating. The rampage detection device according to claim 1, wherein: A storage unit for storing a plurality of images obtained by imaging images in an elevator car by date and time by an imaging device;
Means for capturing a current image by imaging the interior of the elevator car with the imaging device;
Means for extracting a partial area that changes with time from among the partial areas in the current video;
Means for selecting a video in the storage unit corresponding to a date and time as a reference video for the current video;
Means for extracting a partial area that changes with time from among the partial areas in the selected reference image;
Means for comparing the partial region extracted from the current video with the partial region extracted from the reference region;
Means for outputting a partial area that changes with time in the current video,
This output means suppresses the output of the partial area which shows the same change as a result of comparison, The rampage detection apparatus characterized by the above-mentioned. Computer
A plurality of videos obtained by imaging images in the elevator car by date and time by the imaging device are stored in the storage unit,
The imaging device images the inside of the elevator car and stores the current video in the storage unit,
Of the partial areas in this current video, extract the partial areas that change over time,
For the current video, select the video in the storage unit corresponding to the date and time as a reference video,
Extract the partial area that changes over time from the selected partial areas in the reference video,
Comparing the partial area extracted from the current video with the partial area extracted from the reference area;
Configured to output a partial region that changes with time in the current video,
As a result of the comparison, a rampage detection method is characterized in that the output of partial areas showing the same change is suppressed.

Images