JP2011105415A - Elevator system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To overcome the following problem: image recognition by a conventional technique does not detect a flaw or dirtiness in a car nor detect a flaw or dirtiness in the car at an appropriate timing. <P>SOLUTION: The elevator system includes: a lighting system; a monitoring camera; an image recognition device for storing a first photographed image photographed by the monitoring camera in a storage device while the lighting system illuminates the inside of the car; a flaw/dirtiness detection device which compares the first photographed image with a second photographed image previously photographed and which detects a flaw or dirtiness in the car; and a speech recognizer for discontinuing storage into the storage device when a microphone obtains a previously designated voice. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

The present invention relates to an elevator system that detects scratches or dirt in a car from an image taken by a surveillance camera in the elevator car.

As a technology related to monitoring with conventional surveillance cameras, the image taken by the surveillance camera installed in the car is recognized and analyzed to detect the presence or absence of passengers and luggage, and automatically adjust the elevator car load detector. The technique which performs is disclosed (patent document 1).

JP2008-137751

However, image recognition in the prior art relates to automatic adjustment of a load detector that detects the load of a car by analyzing the image in the car and recognizing the presence or absence of a passenger or luggage. It is not intended to detect scratches or dirt on the car, and the load on the car is misrecognized even when the floor or wall surface in the car is replaced and replaced, or when the mover is cured to prevent injury. For this reason, scratches or dirt in the car could not be detected at an appropriate timing.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and includes an illumination device provided in a car, a monitoring camera provided in the car and photographing all or a part of the car as a subject, and an illumination device. When the interior of the car is illuminated, an image recognition device that stores a first captured image captured by the surveillance camera in a storage device, a first captured image stored in the image recognition device, and a first A scratch / dirt detection device for detecting scratches or dirt in the car and a microphone provided in the car to obtain a sound designated in advance is compared with a second photographed image corresponding to the photographed image. And a voice recognition device that stops storing the first captured image in the storage device by the image recognition device.

The present invention further includes a moving object detection device that detects whether or not a moving object is present in the cage. When the moving object detection device determines that there is no moving object in the cage, the image recognition device It is an elevator system which memorize | stores a 1st picked-up image in the said memory | storage device.

In addition, the moving object detection device of the present invention is an elevator system including a scale that is provided between the car and the car frame and calculates the weight of the car.
In addition, the moving object detection device of the present invention is an elevator system including a scale that calculates the amount of change in the weight of the car based on the tension applied to the main rope that suspends the car.

The present invention further includes a door closing detection device for detecting whether or not the car door is closed. When the door closing detection device determines that the car door is closed, the image recognition device performs the first operation. It is an elevator system which memorize | stores a picked-up image in a memory | storage device.

According to the present invention, it is possible to detect scratches or dirt in the car, and in addition, it is possible to prevent misrecognition due to the replacement of the floor and wall surfaces in the car and the injury prevention curing by the mover. Scratches or dirt in the car can be detected at an appropriate time.

Embodiment 1
FIG. 1 is a block diagram showing an elevator system according to the present invention. In FIG. 1, 1 is an elevator car, 2 is installed in the car, and a surveillance camera for the main purpose of preventing crime in the car, 3 is installed in the car, and obtains sound in the car to generate sound. A microphone for converting the signal into a signal, 4 is a voice recognition device for recognizing a specific word from the electrical signal of the sound collected from the microphone 3, 5 is an illuminating device provided in the car to illuminate the car, and 6 is an animal in the car. A moving object detection device that detects whether or not a body exists, 7 is an elevator control device that controls the entire elevator, such as the elevator car 1, 8 is an image recognition device that recognizes an image taken by a surveillance camera, and 9 is a car interior. A flaw / dirt detection device for detecting flaws or dirt, 10 is a monitoring device for notifying a manager that there is a flaw or dirt in the car, and 11 is a manager room of the building where the elevator is provided.

Since the surveillance camera 2 uses an existing surveillance camera whose main purpose is crime prevention, it is not necessary to install a dedicated camera separately, and the cost is reduced. Further, since the microphone 3 uses an existing microphone for communicating with the manager in the manager room 11 from within the car, it is not necessary to install a dedicated microphone separately, and the cost is reduced. The voice recognition device 4 is described here as being in the car, but the present invention is not limited to this and may be installed in the manager room 11. The lighting device 5 adjusts the lighting in the car 1 according to an instruction from the elevator control device 7. The moving object detection device 6 detects whether or not the car 1 has a moving object such as a human being or an animal, and the car is detected by a scale (not shown) that measures the tension of the main rope on which the car 1 is suspended. It may be detected on the basis of the mass change in the sensor, or may be detected by image recognition from the image information of the monitoring camera 2. The elevator control device 7 is connected directly or indirectly to the surveillance camera 2, the voice recognition device 4, the illumination device 5, and the moving object detection device 6 to control them. The elevator control device 7 may be installed in a machine room (not shown) or in an elevator hoistway.

The image recognition device 8 is described here as being in the manager room 11, but is not limited thereto, and may be installed in a central monitoring center (not shown). The scratch / dirt detection device 9 is provided in the image recognition device 8 and compares the captured image 11 captured by the surveillance camera 2 with the captured image 12 captured in advance corresponding to the captured image. It detects dirt. Here, the flaw / dirt detection device 9 is described as being provided independently of the image recognition device 8, but is not limited thereto, and may be installed in the image recognition device 8. The monitoring device 10 will be described as a display as means for presenting information on scratches or dirt in the cage to the manager or the like. However, the monitoring device 10 is not limited to this and may be any device that can transmit information to the manager or the like. But you can.

FIG. 2 is a photographed image photographed by the surveillance camera 2 with all or part of the car as a subject. The photographed image 12 is in a state where, for example, the cage immediately after cleaning is clean, and the photographed image 13 is in a state where scratches or dirt are generated in places in the cage by the user after, for example, the time after cleaning has elapsed. is there.

In the image recognition device 8, the captured image taken by the monitoring camera 2 is stored in the memory when (a) the lighting device 5 is operating, and (b) the moving object detection device 6 puts it in the cage. When it is determined that there is no moving object, and (c) it is determined that the car door is closed by a door closing detection device (not shown) that detects whether the car door is closed. is there. In addition, when the cleaner speaks, for example, “stop image recognition”, which is designated in advance, in order to prevent erroneous recognition due to the replacement of the floor and wall surfaces in the car and the injury prevention curing by the mover, the microphone 3 The voice is collected, and the speech recognition device 4 recognizes the utterance “stop image recognition”, and the elevator control device 7 issues an image recognition command to the image recognition device 8 to stop storing the captured image in the memory. The storage in the memory is interrupted by the device 8. That is, (d) when the microphone 3 does not obtain a predesignated sound, the photographed image is stored in the memory.

Note that, here, it has been described that the captured image is stored in the memory when all of the above requirements (a), (b), (c), and (d) are satisfied, but the present invention is not limited to this. If only the requirement () is satisfied, the captured image may be stored in the memory.

When there is no stop command, a signal of a photographed image obtained by photographing the interior of the elevator car with the monitoring camera 2 attached in the elevator car 1 passes through the elevator control device 7 and is an image recognition device in the manager room 11. 8 is sent.

Here, it may be based on the captured image to determine whether or not the lighting device 5 in the car is operating, or a brightness detection device may be separately provided in the car (not shown). The determination of whether a moving object is present can be based on a photographed image, or the change in the weight of the car can be detected by a scale (not shown) that calculates the tension of the main rope that suspends the car. The basis may be.

The scratch / dirt detection device 9 compares the previously recorded image 12 with the currently recorded image 13 when a recorded image exists, and if there is a difference described later between the images, scratches or dirt are generated in the car. The monitoring device 10 in the administrator room 11 is notified. In the monitoring device 10, a message prompting the administrator to repair or clean is displayed on the screen by a display or by a voice by a speaker.

The previous photographed image is preferably in a clean state after the cleaning, and the previous photographed image 12 is obtained when the cleaner speaks, for example, “image recognition”, the microphone 3 collects this sound, The speech recognition device 4 recognizes the utterance “image recognition” and transmits a signal to that effect to the elevator control device 7, and the elevator control device 7 sends the captured image 12 taken by the monitoring camera 2 to the image recognition device 8. May be stored in the memory.

The scratch / dirt detection device 9 compares the image 13 photographed by the surveillance camera 2 with the previously recorded image 12 when it is determined by the scale that the moving object is not present in the cage, and the graffiti 14 If there are differences such as scratch 15, liquid 16, and dust 17, it is determined that scratches or dirt have occurred in the elevator car. The detection of the difference is an unmanned state, and the image is recorded in a state where the luminance and color change of the squares of the photographed image are lost, and usually there is a difference from the image 12 stored one time before. If there is a difference in brightness or color between all squares of the image, it is determined that scratches or dirt have occurred in the elevator car.

Next, the operation of this elevator system will be described with reference to the flowcharts of FIGS. In FIG. 4, before the work is performed in the car, the person in charge issues a keyword for temporarily stopping image detection by voice, and the voice recognition device 4 attached to the car recognizes the keyword for temporarily stopping image detection. FIG.

In S21, when the lighting device 5 in the car is not operated, when the moving object detection device 6 determines that the moving object is present in the car, or when the door is opened by the door closing detection device. If it is determined, S21 is repeated.

In S21, the lighting device 5 in the car is operating, the moving object detection device 6 determines that there is no moving object in the car, and the door detection device determines that the door is in a closed state. In this case, in S22, the captured image captured by the monitoring camera 2 is stored in the memory in the image recognition device 8.

In S <b> 23, the image recognition device 8 confirms whether the captured image 12 that has already been stored exists in the memory.

In S23, if there is no stored image 12 already stored, the process is repeated from S21. If the photographed image 12 already stored in S23 exists, the scratch / dirt detection device 9 compares the stored photographed image 12 with the photographed image 13 stored this time in S24. Determine if there is any.

If it is determined in S25 that there is no difference between the previously stored captured image 12 and the currently stored captured image 13, the process is repeated from S21. When it is determined that there is a difference between the captured image 12 stored last time and the captured image 13 stored this time in S25, it is determined that scratches or dirt have occurred in the car, and the detection of scratches or dirt is detected in the monitoring apparatus 10 in S26. Notice. The monitoring device 10 that has received the notification notifies the manager of the occurrence of scratches or dirt in the car by screen display or voice.

In this way, the lighting device 5 in the car is operating, there is no moving object in the car, and the door is in a closed state, the existing stored image 12 and the currently stored image 13 are compared, It can detect scratches or dirt in the car and inform the manager that cleaning or repair is necessary, and if a scratch or dirt occurs in the car, the manager etc. will immediately go to the car. Can be cleaned.

In addition, the existing stored image 12 and the currently stored image 13 were compared with each other due to the replacement and replacement of the floor and wall surfaces of the car room and the injury prevention by the moving contractor, so that scratches or dirt were generated in the car. May be falsely detected. In order to prevent this erroneous recognition, a keyword that pauses the operation of storing the captured image in the memory by voice (for example, “stop image recognition”) is uttered before performing the operation in the car. When the voice recognition device 4 installed in the car recognizes a keyword that is a predetermined voice, the voice recognition device 4 notifies the image recognition device 8 via the elevator control device 7. The image recognition device 8 processes the flowchart of FIG. 4 when receiving the notification.

In S31, the captured image for determination recorded so far is discarded and set in an initial state, and in S32, storage of the captured image in the memory of the image recognition apparatus 8 is set in a suspended state. The function resumption from the paused state is notified to the image recognition device 8 by issuing a resumption keyword (for example, “image recognition”) in the car and being recognized by the voice recognition device 4 attached in the car, or Then, a restart notification is transmitted to the image recognition device 8 by an operation from the monitoring device 10. When the image recognition apparatus 8 receives the restart notification, the process of the flowchart of FIG. 3 is performed from the beginning.

Thus, when the memory of the image of the image recognition apparatus 8 is clearly erroneously recognized, the function is temporarily stopped by making a sound in the car in advance, and can be resumed by making the sound in the car from the stopped state. By doing so, it is possible to avoid misrecognition with a simple procedure. For this reason, it is possible to prevent the manager from going to the car to repair or clean the inside of the car based on the erroneous recognition.

Embodiment 2. FIG.
In the present embodiment, the flowchart of FIG. 5 is processed by the image recognition device 8 of the elevator image monitoring system of FIG. After the processing of S21 to S26 is performed by the image recognition apparatus 8 according to the first embodiment and scratches or dirt are detected, the state is set to a scratch or dirt detection alarm state, and the flowchart of FIG. 5 is performed.

First, in S41, the state in the car is determined. If it is determined in S41 that the lighting device 5 in the car is OFF, or the moving object detection device 6 determines that the moving object is present or is in the door open state, no processing is performed and the process is repeated from S41.

In S41, when the lighting device 5 in the car is ON, and the moving object detection device 6 determines that the moving object is not present and is in the door-closed state, the captured image in the car is stored in S42. In S43, the photographed image before and after the detection of scratches or dirt is compared with the photographed image stored this time, and in S44, it is determined whether the photographed image before the detection of scratches or dirt is close to the photographed image after the detection of scratches or dirt.

If the photographed image stored this time in S44 is close to the photographed image after detection of scratches or dirt, the process is repeated from S41.

If the photographed image stored this time in S44 is close to the photographed image before the detection of scratches or dirt, the monitoring device 10 is notified of the restoration, the scratch or dirt detection alarm state is canceled, and the flowchart of FIG. carry out. The monitoring device 10 stops the flaw or dirt detection screen display or voice notification for the manager.

In this way, when the elevator user notices a fallen object and comes to pick it up by automatically detecting that it has been restored, or if it is immediately repairing a scratch or cleaning dirt, the dirt detection status Can be automatically released, and the manager can recognize that there is no need for cleaning or repair.

1 is a block diagram of an elevator image monitoring system according to a first embodiment of the present application. FIG. : It is an example of the image in a car recognized with the elevator image monitoring system concerning Embodiment 1 of this application. : It is a flowchart in which the image recognition apparatus of the elevator image monitoring system concerning Embodiment 1 of this application detects a crack or dirt. : It is a flowchart when the image recognition apparatus of the elevator image monitoring system concerning Embodiment 1 of this application receives a notice of a pause by voice recognition for a crack or dirt detection. : It is a flowchart in which the image recognition apparatus of the elevator image monitoring system of Embodiment 2 of the present application detects the recovery of scratches or dirt.

1: elevator car, 2: car monitoring camera, 3: microphone, 4: car voice recognition device, 5: lighting device, 6: moving object detection device, 7: elevator control device, 8: image recognition device, 9: Scratch / dirt detection device, 10: monitoring device, 11: manager's room, 12: in-car image (scratch or dirt not detected), 13: in-car image (scratch or dirt detection), 14: graffiti, 15: scratch, 16: Liquid, 17: Garbage

Claims (4)

A lighting device provided in the cage;
A surveillance camera that is provided in the car and shoots all or part of the car as a subject;
An image recognition device that stores a first captured image captured by the monitoring camera in a storage device when the illumination device illuminates the car;
The first photographed image stored in the image recognition device is compared with the second photographed image previously photographed corresponding to the first photographed image, and a scratch or dirt in the car is detected. A dirt detector,
A voice recognition device for stopping the storage of the first photographed image in the storage device by the image recognition device when a microphone provided in the car obtains a predetermined voice;
An elevator system characterized by comprising: A moving body detecting device for detecting whether a moving body is present in the cage;
2. The elevator according to claim 1, wherein the first captured image is stored in the storage device by the image recognition device when it is determined by the moving object detection device that no moving object is present in the car. system. The elevator system according to claim 2, wherein the moving object detection device includes a scale that calculates a change amount of the weight of the car based on a tension applied to a main rope that suspends the car. Further comprising a door closing detection device for detecting whether the car door is closed,
2. The elevator system according to claim 1, wherein the first photographed image is stored in the storage device by the image recognition device when it is determined by the door closing detection device that the door of the car is closed. .

Images