JP2875429B2 - Anomaly detection system for passengers - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for detecting a passenger abnormality of a man-conveyor which detects an abnormality of a passenger of a man-conveyor such as an escalator or a moving sidewalk from image data of the passenger.

[0002]

2. Description of the Related Art First, if the technical background of this type of apparatus is described, when a passenger of a man conveyor such as an escalator or a moving walkway falls near the exit of the man conveyor, the passenger on the succeeding man conveyor will get off. The mouth is blocked and stayed, causing a fall accident. Further, even when the passenger enters the man conveyor in a reverse direction against the moving direction of the man conveyor, the passengers may collide with each other and cause the same accident.

[0003] For this reason, the fall of the passenger, the reverse approach, etc. are detected at an early stage, and measures such as notifying the observer or stopping the operation of the man conveyor are taken. Need to be prevented. In order to respond to such a need, conventionally, when the number of passengers near the exit increases and the passengers easily fall, the speed of the man conveyor is reduced, and the risk of falling is eliminated as much as possible.

FIG. 11 shows, for example, Japanese Patent Application Laid-Open No. 47-25883.
FIG. 1 is a cross-sectional view of a mechanism section in a conventional passenger abnormality detection apparatus for a man conveyor shown in Japanese Patent Application Laid-Open Publication No. HEI 10-152139. FIG. 12 is a circuit diagram of an operation control unit for switching the driving of the man conveyor from the high-speed induction motor to the low-speed induction motor in the conventional apparatus. In FIG. 11, reference numeral 101 denotes a driving unit of a man conveyor,
As a truss provided with a drive mechanism or the like, a drive unit provided on the truss 101, or a drive mechanism 102, a high-speed induction motor and a low-speed induction motor are incorporated on the same shaft, and an energy storage wheel is mounted on this shaft. A drive unit (not shown) is mounted, 103 is a drive chain gear fixed to the shaft, 104 is a drive chain wound around the drive chain gear 103, 105 is a drive chain also wound around the drive chain 104 A main chain gear 106 driven by the machine 102 is a pallet chain gear fixed coaxially with the main chain gear 105, and a pallet chain 107 is provided between the pallet chain gear (not shown) and one of the pallet chain gears provided in the longitudinal direction. It is wound endlessly.

[0005] 108a-108n are pallet chains 107
Roller shafts 109a to 109n connected at equal pitch to
Is a roller pivotally attached to both ends of each of the roller shafts 108a to 108n, and 110 is a moving tread plate mounted over the roller shaft 108a to 108n, so that passengers of a man conveyor can ride on the upper surface of the moving tread plate 110. It is configured.
111 is a rail for guiding the rollers 109a to 109n;
Reference numeral 112 denotes a switch for detecting a load, which is provided on the back surface of the rail 111 near the exit and is activated by the pressing force of the radius of the rail 111 when the load on the rail 111 by the passenger on the moving tread plate 110 exceeds a predetermined value. I do. Reference numeral 113 denotes a floor plate constituting the exit of the man conveyor, and reference numeral 114 denotes a balustrade of the man conveyor.

[0006] In FIG. 12, 119a ₁ ~119a ₃ power terminals R, S, contacts for controlling the energization of the three-phase AC power supply connected to the T, 120a ₁ ~120a ₃ is high-speed induction motor three-phase AC power source A contact for controlling the energization of 115, 121a
Numerals _{1 to} 121a ₃ denote a three-phase AC power source and a low-speed induction motor 116.
This is a contact point for controlling the energization. Next, 11 which is connected between the plus (+) and minus (-) DC power supply lines.
Reference numeral 7 denotes an operation push button, 118 denotes a stop push button, and 119 denotes an operation relay which is energized when an excitation current is received via the operation push button 117. The energization of the operation relay 119 causes a contact 119a ₁ To 119a ₃ and 119b are closed.
124a, 124c are contacts of the relays for switching operation to be described later, 120 is a high speed operation relay to close the contacts 120a ₁ ~120a ₃ biases receives the excitation current through the contacts 124c, 121 via the contact 124a the contact 121a ₁ ~121a ₃ biased when receiving an exciting current is low speed operation relay to be closed.

Reference numeral 122 denotes a timed relay, and the switch 11
2 is a timed relay that is energized after a lapse of a predetermined time when DC voltage is supplied via the power supply 2 and is immediately deenergized when the power is cut off. 122a is a normally open contact of the timed relay 122, 122b
Is a normally closed contact of the timed relay 122, and 123 is a normally open contact 12.
A time relay that performs the same operation as the time relay 122 by closing 2a, an operation switching relay 124 that switches between high-speed operation and low-speed operation of the man conveyor, and a high-speed operation relay 120 that is activated by the operation switching relay 124 The energizing operation is switched to the low-speed operation relay 121. 12
Reference numeral 5 denotes a time relay that performs the same operation as the time relay 122. Reference numeral 126 denotes a shift indicator light for notifying the passenger that the traveling speed of the conveyer is changed, and 127 an alarm device for notifying the speed change by an alarm sound. These notification means include a contact point 119c of a driving relay 119 and a time limit. Relay 123, 1
An operating current flows through each of the 25 contacts 123a and 125a.

[0008] Next, the operation of the conventional passenger abnormality detection apparatus for a man conveyor according to the above configuration will be described. Now, when a man conveyor, for example, a moving sidewalk, is stopped, the switch 112 is not depressed, and the switch contact is turned on to the b side. Therefore, the timed relay 122 is in an energized state, closes its own normally open contact 122a, and energizes the timed relay 123. As a result, the normally closed contact 123a of the timed relay 123 is open,
26, the alarm 127 does not operate.

When the operation push button 117 is depressed while the switch 112 is turned on as described above, the operation relay 119 is energized, and the normally open contact 119b is closed, so that the plus-stop button-normal Open contact 119b-operation relay 119
Operation relay 1 with a closed loop such as minus
19 self-holds the biasing operation. A contact 119 connected between each AC power supply line by the energization of the operation relay 119
a _{1 to} 119a ₃ are closed, and the three-phase AC power supply is turned on. Further, when the high-speed operation relay 120 receives an excitation current and is energized through the stop button 118, the normally-open contact 119b, and the normally-closed contact 124c, the high-speed operation relay 120, which is connected to the power input side of the high-speed induction motor 115, is normally opened. the contact 120a ₁ ~120a ₃ to start the three-phase AC power source and closed by energizing the high-speed induction motor 115. As a result, the tread 110 moves in the direction of arrow A, and the moving sidewalk performs high-speed driving.

When the passengers on the tread 110 are crowded particularly near the landing and a load exceeding a predetermined level is applied to the tread 110, the switch 112 operates to switch the contact point to the a side. As a result, the timed relays 122 and 123 are immediately deenergized due to the power interruption, and the normally closed contacts 122b and 123a of the timed relays 122 and 123 return to the closed state. Normally closed contact 12
By the return of 2b, the operation switching relay 124 receives the excitation current via the switch 112 and the normally closed contact 122b to be energized, and closes its own normally open contact 124b to energize the timed relay 125. However, even if the timed relay 125 is energized, its own normally closed contact 125a does not become open immediately and opens after a certain period of time.

Therefore, until the normally closed contact 125a is opened, a drive current flows through the normally open contact 119a and the normally closed contact 125a to the speed change indicator lamp 126 and the alarm 127, and the sounding operation and the sounding of the alarm sound are performed. Is started. As a result,
Passengers are notified that the speed of the moving sidewalk will change from low to high. Further, the self-closed contact 124c is opened by the bias of the operation switching relay 124,
When the normally open contact 124a is closed, the high-speed relay 120 is de-energized, and the low-speed relay 121 is energized. Thereby the contact 120a ₁ ~120a ₃ is opened, since the contact 121a ₁ ~121a ₃ by Hen' closing, high-speed induction motor 115 is split off from the three-phase AC power source, the low-speed induction motor 116 three-phase AC Power supply contacts 121a _1-
It is connected via 121a _3.

As described above, even if the driving motor for the moving sidewalk is switched from the high-speed induction motor 115 to the low-speed induction motor 116, the inertia moment of the energy storage wheels mounted on the shafts of the induction motors 115 and 116 is large. Therefore, the moving sidewalk does not decelerate rapidly, but gradually decelerates to a constant speed. In the process of reaching the constant speed, the timed relay 125 is energized after a given time elapses, and the normally closed contact 125a
Is released, the shift indicator lamp 126 is turned off, and the alarm 127 stops sounding. Since the sidewalk moving in this way continues the low-speed operation, the passengers can safely get off the floor one by one.
13 can be descended.

At low speed operation, the tread 11 near the landing
When the number of passengers on zero decreases and becomes equal to or less than the predetermined load, the switch 112 returns, and the contact switches again to the b side as shown in FIG. As a result, the plus relay 112, the timed relay 122, and the minus closed circuit make the timed relay 1
When a predetermined time (for example, 30 seconds) elapses after the excitation current flows through the switch 22, the normally-closed contact 122b is opened and the operation switching relay 124 is de-energized. Operation switching relay 1
The normally open contact 124b is opened by the deenergization of 24, and the normally closed contact 125a of the timed relay 125 is restored when the power of the timed relay 125 is turned off and deenergized.

At this time, the normally open contact 12 of the timed relay 122
2a has not yet been closed, so that the normally open contact 122
a timed relay 123 connected in series to
The normally closed contact 123a connected in series to the shift indicator light 126 and the alarm 127 is closed, and the power supply is closed for 30 seconds until the normally closed contact 123a is opened.
For a second, the shift indicator lamp 126 is turned on, and the alarm 127 is sounded to notify the passenger that the moving sidewalk is switched from low speed to high speed driving.

Further, when the operation switching relay 124 is deenergized, the self-closed contact 124c returns to a reset state.
When the normally open contact 124a is opened, the high speed relay 120 is energized, and the low speed relay 121 is deenergized.
Thereby the contact 120a ₁ ~120a ₃ is closed, the contacts 1
21a ₁ for ～121A ₃ is opened, the low-speed induction motor 115 is split off from the three-phase AC power source, the high-speed induction motor 116 is connected through a contact 121a ₁ ~121a ₃ into three-phase AC power source.

Even if the driving motor for the moving sidewalk is switched from the low-speed induction motor 116 to the high-speed induction motor 115 as described above, the energy storage wheels are mounted on the shafts of the induction motors 115 and 116. Therefore, the sidewalk where the moment of inertia moves greatly does not decelerate rapidly, but gradually decelerates to a constant speed. When a certain period of time elapses in the process of reaching the constant speed, the time relay 123 is energized, the normally closed contact 123a is opened, the shift indicator lamp 126 is turned off, and the alarm 127 stops sounding.

[0017]

As described above, the conventional passenger abnormality detection apparatus for a man-conveyor switches the operation of the man-conveyor to low-speed operation when the amount of passengers in the vicinity of the exit exceeds a predetermined amount. In order to prevent passengers from falling down in the vicinity, for example, passengers moving in the direction of movement or in the opposite direction while looking aside on the pussy conveyor and disturbing the normal flow of passengers in the man conveyor There is a problem that it is not possible to cope with abnormalities such as falling down.

Further, reducing the driving speed when the passengers are crowded has a problem that the transportation efficiency of the passengers is reduced and further congestion is caused.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and monitors the flow of passengers in a pussy bear and issues an alarm to a supervisor when a fall of the passenger is detected, or controls the speed of the pussy bear. It is an object of the present invention to obtain a passenger abnormality detecting device for a man conveyor that can prevent the accident from spreading.

[0020]

According to a first aspect of the present invention, there is provided an apparatus for detecting an abnormality of a passenger on a conveyer, comprising: an image pickup unit for picking up an image of a passenger on the conveyer to obtain image data; and the image determined by a moving direction of the conveyer. Abnormality determining means for detecting a disturbance in the moving direction of the data and determining an abnormality in the passenger flow.

According to a second aspect of the present invention, there is provided an apparatus for detecting an abnormality of a passenger on a passenger conveyor, comprising: an imaging unit for capturing an image of a passenger on the passenger conveyor to obtain image data; and after extracting a characteristic image such as a contour of the passenger from the image data. A flow observing means for arranging the moving direction of the characteristic image for each transition of time, observing the flow of the passenger accompanying the movement of the conveyor, and extracting a turbulence of the flow from the observed flow of the passenger. And a disturbance extracting means.

According to a third aspect of the present invention, there is provided an apparatus for detecting an abnormality of a passenger on a passenger conveyor, comprising: an imaging unit for capturing an image of a passenger on the passenger conveyor to obtain image data; and after extracting a characteristic image such as a contour of the passenger from the image data. A flow observation means for arranging the moving direction of the feature image for each transition of time and observing the flow of the passenger accompanying the movement of the conveyor, and a flow direction of the passenger observed by the flow observing means. A flow direction extraction unit that extracts the inclination direction value, and a comparison between the inclination direction value extracted by the flow direction extraction unit and the inclination direction value when the passenger flow is steady. And a flow direction comparing means for determining the turbulence of the flow of the passenger based on the frequency of extraction of the direction value.

According to a fourth aspect of the present invention, there is provided an apparatus for detecting an abnormality of a passenger on a man conveyor, which is configured to image the passengers on the man conveyor and the passengers near the entrance / exit of the man conveyor, and obtain image data of each passenger; After extracting characteristic images such as contours of each passenger from the data, the moving directions of the characteristic images are arranged for each transition of time, and the first and second observations of the flow of each passenger accompanying the movement of the man conveyor are performed. (2) flow observing means, and first and second extraction means for extracting the direction of the flow of the passenger on the conveyer and the flow of the passenger near the entrance / exit of the conveyer, which are observed by the first and second flow observing means, respectively, by using the inclination direction value.
Comparing the inclination direction value extracted by both the first and second flow direction extraction means with the inclination direction value when the passenger flow is steady; And a flow direction comparing means for determining the turbulence of the flow of the passenger based on the value extraction frequency.

[0024]

According to the first aspect of the present invention, a passenger abnormality detecting apparatus for a passenger conveyor determines an abnormality in the passenger flow based on image data obtained by imaging passengers of the passenger conveyor and detects occurrence of an accident.

The passenger conveyor anomaly detection device according to the second aspect of the present invention extracts the passenger flow of the passenger conveyor over time and detects an occurrence of disturbance in the passenger conveyor. Determine that it has occurred.

According to the third aspect of the present invention, the passenger abnormality detecting apparatus for the passenger conveyor captures the moving direction of the passenger of the passenger conveyor with time as a two-dimensional image, and uses the image to correspond to the moving direction of the passenger. When the occurrence frequency of each inclination direction value becomes irregular, it is determined that the flow of the passenger is disturbed and an accident has occurred.

According to a fourth aspect of the present invention, there is provided an apparatus for detecting a passenger abnormality in a passenger conveyor, which determines whether or not the passenger flow has been disturbed both at the passenger on the passenger conveyor and at the entrance / exit of the passenger conveyor based on the frequency of occurrence of each inclination direction value. Since the judgment is made, the occurrence of the accident can be monitored with a wider field of view.

[0028]

【Example】

Embodiment 1 FIG. FIG. 1 is a block diagram showing an embodiment of a configuration of a passenger abnormality detection apparatus for a conveyor according to the present invention.
The present invention also includes the imaging means, ie, the ITV camera 1 and the abnormality determination means 10, which constitute the invention of the first aspect. In the figure, reference numeral 1 denotes an ITV camera installed on a ceiling near an entrance / exit of an escalator, which is a kind of a man conveyor, and outputs image data indicating a passenger flow. 2 is ITV camera 1
The first area setting means 3 previously sets an area where the steps of the escalator is taken as the first area in the image data output from the escalator, and the area near the entrance of the escalator is included in the image data output from the ITV camera 1. A second area setting means for presetting a handed area as a second area;
A feature extraction unit 5 for processing image data in the region and the image data in the second region to extract features such as a passenger contour is a feature projection unit, and a feature extraction result extracted by the feature extraction unit 4 is used as a first feature extraction result. And in the direction perpendicular to the direction of movement of the passenger in the second area, and a one-dimensional image of the movement direction axis is generated for each area.

Reference numeral 6 denotes a moving direction-time axis image generating means. The one-dimensional images of the moving direction axis obtained by the characteristic projecting means 5 are arranged in chronological order to form a two-dimensional image of the moving direction-time axis. Numeral 7 denotes a tilt extracting unit which calculates a two-dimensional tilt of a line segment for each pixel by performing a process such as a local filter on the two-dimensional image generated by the moving direction-time axis image generating unit 6. To generate a tilt direction image having a tilt direction value. Numeral 8 denotes a tilt direction frequency calculating means, which displays the tilt direction images of the first and second regions calculated by the tilt extracting means 7 along the moving direction axis at each time and tilt direction value. Calculate the frequency.

Reference numeral 9 denotes frequency evaluation means, and the first and second frequency evaluation means
Tilt direction frequency calculating means 8 for each time of region and for each tilt direction
Is processed to calculate how much the appearance frequency of the current inclination direction value differs from the appearance frequency of the inclination direction value when the flow of escalator passengers is steady. 1
The reference numeral 0 denotes an abnormal signal output means. If the frequency of occurrence of the inclination direction value is significantly different from that in the steady state based on the output of the frequency evaluation means 9, it is determined that an abnormality has occurred in the passenger. 1
Reference numeral 1 denotes an abnormal signal output unit that outputs an abnormal signal when a passenger abnormality is determined.

FIG. 5 is a diagram showing a first area and a second area set by the first area setting means 2 and the second area setting means 3. In the figure, 12 is an input image obtained by photographing with the ITV camera 1, 13 is a first area shown in the input image, 14 is a second area also shown in the input image, and 15 is a first area Escalator steps shown inside,
Reference numeral 16 denotes an entrance of the escalator shown in the second area. Reference numeral 17 denotes a passenger at a certain time T, and reference numerals 17a to 17l denote passengers from a time T + 1 to a time T + 12, which indicate the moving positions of the passengers based on the transition of time.

FIG. 6 is a diagram for explaining the operation of a series of means from the feature extracting means 4 to the moving direction-time axis image generating means 6. In the figure, 18a to 18c represent times T to T + 2
Are the passengers shown in the input images 18a to 18c, respectively, and are moving in the direction of arrow B in the figure. 21a to 21c are characteristic images extracted by the characteristic extracting means 4 from the input images 18a to 18c, and 22a to 22c and 23a are passengers 19.
a to 19c and 20a. 24a
24c are one-dimensional images in the moving direction axis, which are the result of performing feature projection on the feature images 21a to 21c, and 25a to 25c and 26a in the one-dimensional images 24a to 24c indicate locations where the features of the passenger exist. Represent.

Numeral 27 denotes a moving direction-time axis image generated by the moving direction-time axis image generating means 6 based on the places 25a to 25c and 26a where the features changing with time exist. Reference numeral 28 denotes a movement track of the passengers 19a to 19c, and reference numeral 29 denotes a movement track of the passenger 20a.

FIG. 7 is a conceptual diagram showing an example of a moving direction-time axis image. In the figure, 30 is the moving direction-time axis image obtained from the first area 13, 31 is the moving direction-time axis image obtained from the second area 14, and 32 is the movement of the step 15 existing in the first area 13. , A movement trajectory representing the movement of the passenger in the first area 13, a movement trajectory representing the normal movement of the passenger in the second area 14, 35
Is a movement trajectory representing an abnormal movement such as a fall of the passenger in the second area 14.

FIG. 8 is a diagram showing a direction mask which is a local filter for extracting the inclination direction from the moving direction-time axis image. In the figure, reference numerals 36a to 36h denote 9 × 9 masks for extracting the inclination direction values 1 to 8, respectively. FIG. 9 is a conceptual diagram for explaining the operation from the inclination extracting means 7 to the inclination direction frequency calculating means 8. In the figure, 31
From the movement direction-time axis image obtained from the second area 14,
Reference numeral 34 denotes a movement trajectory representing a normal movement of the passenger in the second area 14, 35 denotes a movement trajectory representing an abnormal movement of the passenger in the second area 14 such as a fall, and 37 denotes a movement of the second area 14. The inclination direction images 38a to 38h obtained from the direction-time axis image 31 by the inclination extraction means 7 are the inclination direction images 37.
Is the frequency of appearance of each of the inclination direction values 1 to 8 obtained by the inclination direction frequency calculation means 8 from FIG.

FIG. 7 is a diagram for explaining detection of reverse entry of a passenger. In the figure, 30 is a moving direction-time axis image obtained from the first area 13, 32 is a moving path representing the movement of the steps 15, 33 is a moving path representing the movement of a passenger in the first area 13, and 39 is a moving path. This is a movement trajectory representing the movement of the passenger when the passenger enters the area 13 in a direction opposite to the direction of the other passengers.

Next, the operation of this embodiment will be described. First, an area setting operation will be described with reference to FIGS. I
From the input image 12 captured from the TV camera 1, the first
An image of a portion corresponding to the first area 13 and the second area 14 is extracted by the area setting means 2 and the second area setting means 3. At this time, the first area 13 is an area where the step 15 of the escalator is photographed, and is stored in the first area setting means 2 as an area where the passenger is moving from top to bottom in the direction of arrow B. Further, the second area 14 is an area in which the entrance of the escalator is taken in the input image, and the passenger is stored in the second area setting means 3 as an area moving from right to left in the direction of arrow C. I have. That is, at this stage, the input image 12 is divided according to the moving direction of the passenger.

Next, the processing operation up to the generation of the moving direction-time axis image will be described with reference to FIGS. 1, 6, and 7.
Incidentally, regarding the processing during this time, the first area 13 and the second area 1
In this case, the first area 1
Only the third image will be described. First, feature extraction means 4
Is an image 18a at time T of the first area 13 extracted by the first area setting means 2 (the passenger 19a is indicated by an arrow B in the figure).
) Is extracted by edge extraction processing as in, for example, first-order differentiation processing (Oma Makoto, "Image Recognition", Corona Co., p. 46), to extract the characteristics of the passenger, and the characteristic image 2
Obtain 1a. At this time, the edge 22a is extracted as a feature of the passenger 19a.

Next, this feature image 21a is projected by the feature projection means 5 in a direction perpendicular to the moving direction of the passenger to produce a one-dimensional image 24a in the moving direction axis (the width in the horizontal direction is one pixel, Is the number of pixels on the Y axis ) . At this time, the place where the passenger characteristic 22a exists is represented by black 25a, and the place where the passenger characteristic 22a does not exist is represented by white. The above processing is performed on the input image 18b at time T + 1 and the input image 18 at time T + 2.
Similar processing is performed for c to obtain the one-dimensional images 24b and 24c in the movement direction axis at each time. These one-dimensional images 24a to 24c are arranged in time order by the moving direction-time axis image generating means 6, and a moving direction-time axis image 27 is obtained.
In the figure, the horizontal axis represents the time axis, and the vertical axis represents the moving direction axis. Moving direction-time axis image 2 generated in this way
7, the movement of the passengers 19a to 19c is one belt area 2
8, the movement of the passenger 20a is represented by another band area 29.
It is represented by

As a result, the moving direction-time axis image of the first area 13 is as shown by 30 in FIG. 7 , and the moving direction-time axis image of the second area 14 is as shown by 31. Of the first region 13
Movement direction- band region 32 existing above time axis image 30
Is a movement locus representing the movement of the steps 15 of the conveyor, and the movement of the belt area 33 representing the movement of the passenger (since the movement is within the first area, this is the movement of the passenger on the steps 15). It is a trajectory. In addition, the movement direction of the second area 14-the band area 34 in the time axis image 31 also indicates the movement of the passenger (since the movement is in the second area, these movements after the passenger gets on and off at the entrance). Represents. This drawing shows that a plurality of passengers got on and off because there are a plurality of band regions 33 and band regions 34.

Now, when the movement of the passenger is normal, the passenger moves linearly at substantially the same speed in each of the first and second areas. If the passenger falls at the time C (near the entrance), for example, the straight line is lost, and the constant speed and the linearity of the faller are lost. The moving trajectory is disturbed like this. Therefore, by detecting the disturbance of the moving trajectory of the passenger and the area, it is possible to determine whether or not the passenger has an abnormality (in this case, a fall).

Next, a process of calculating the inclination direction frequency, which is a preparation for detecting the disturbance of the movement locus of the passenger, will be described with reference to FIGS. Slope extraction means 7
And the inclination direction frequency calculating means 8, the first area 1
3. Since the image common processing of the second area 14 is performed, only the moving direction-time axis image 31 of the second area 14 will be described here. The tilt extraction means 7 scans the entire screen of the moving direction-time axis image 31 with N × M (for example, 9 × 9) direction masks 36 a to 36 h, and scans each pixel for 8 pixels.
Directional intensities are obtained (for example, the directional intensity
For each of a to 36h, the value of the pixel at the coordinate marked “1” is added, and this value is obtained by dividing by 9). A direction mask indicating the intensity is found, and the direction indicated by the direction mask is set as a tilt direction of the pixel.

As a result, the tilt direction pixel 37 as shown in FIG. 9 is output. In the figure, the direction indicated by the arrow is the inclination direction of each pixel. The tilt direction frequency calculating means 8 scans the tilt direction image 37, counts the number of pixels indicating which tilt direction is present on the moving direction axis at each time, and calculates the tilt direction for each tilt direction. The frequencies 38a to 38h are calculated.
Note that both the inclination direction values 1 and 8 are extracted from the band area 34 indicating the trajectory of the passenger because the inclination direction to be extracted is quantized into eight directions (22 per direction). This is because the entire inclination angle of the band region 34 has an intermediate angle with respect to (.5 degrees). Similarly,
The inclination direction frequency is also obtained from the moving direction-time axis image 30 obtained from the image of the first area 34.

Finally, a description will be given of a process for detecting a disturbance in the trajectory of the passenger based on the inclination direction frequencies 38a to 38h calculated in the above process. When the passenger in the second area 14 has no abnormality (when there is no abnormality in the state near the entrance / exit), the inclination direction frequency (inclination direction frequency with respect to the band area 34) includes a direction value 1 and a direction value 8, There are almost no other tilt directions.

On the other hand, when there is an abnormality (for example, when the passenger falls over near the entrance as described above), the inclination direction frequency (the inclination direction frequency with respect to the band area 35) is different from the direction values 1 and 8. Existing. Therefore, frequency evaluation means 9
For example, the frequency of appearance of the inclination direction value 1 and the frequency of appearance of the inclination direction value 8 are added for each time, while the frequency of appearance of other inclination direction values (for example, in an inclination direction completely different from the inclination direction values 1 and 8). Some tilt direction values 5 and 6 are also temporally added, and the ratio of the addition results of both (for example, the addition result of the tilt direction values 5 and 6 加 算 the addition result of the tilt direction values 1 and 8) is determined to determine abnormality. The means 10 determines that the passenger has an abnormality when the ratio has a value significantly different from the normal state (for example, the ratio increases), and outputs a signal indicating the abnormality of the passenger from the abnormality signal output unit 11. Is output. Also, the first area 1
In the case where there is an abnormality in the passenger No. 3 (for example, when the passenger falls down on the steps 15), the moving direction-time axis image 30
Since the movement trajectory is disturbed, an abnormality can be determined by the same processing.

Embodiment 2 FIG. FIG. 2 is a block diagram showing a configuration of an embodiment of a passenger conveyor passenger abnormality detecting device according to the second aspect of the present invention. In FIG. 1, reference numeral 1A denotes a flow observation unit composed of a feature extraction unit 4, a feature projection unit 5, and a movement direction-time axis image generation unit 6. Reference numeral 1B denotes a turbulence extracting means constituted by the inclination extracting means 7, the inclination direction frequency calculating means 8, the frequency evaluating means 9, and the abnormality judging means 10. With this configuration, the flow of passengers on the conveyor along with the transition of time is extracted, and when it is detected that the flow has been disturbed, it is determined that an accident has occurred between the passengers.

Embodiment 3 FIG. FIG. 3 is a block diagram showing the configuration of an embodiment of a passenger conveyor passenger abnormality detecting apparatus according to the third aspect of the present invention. In the figure, 1A is a flow observation means constituted by a feature extraction means 4, a feature projection means 5, and a movement direction-time axis image generation means 6, and 1C is a direction constituted by a slope extraction means 7 and a inclination direction frequency calculation means 8. The extracting means and 1D are flow direction comparing means constituted by the frequency evaluating means 9 and the abnormality determining means 10. With this configuration, the moving direction of the passenger on the conveyor along with the transition of time is taken as a two-dimensional image, and the moving direction of the passenger is represented by a tilt direction value corresponding to the moving direction from this image, and the frequency of occurrence of each tilt direction value When the vehicle becomes irregular, it is determined that an accident has occurred due to disturbance of the passenger flow.

Embodiment 4 FIG. FIG. 4 is a block diagram showing the configuration of an embodiment of a passenger conveyor passenger abnormality detecting apparatus according to the fourth aspect of the present invention. In the figure, 1A ₁ and 1A ₂ denote _first and second flow observing means, 1C ₁ and 1C ₂ respectively composed of a feature extracting means 4, a feature projecting means 5 and a moving direction-time axis image generating means 6. First and second flow direction extracting means 1D composed of the inclination extracting means 7 and inclination direction frequency calculating means 8 are flow direction comparing means composed of the frequency evaluating means 9 and the abnormality determining means 10. With such a configuration, the occurrence of an accident is monitored from a wider field of view in order to determine whether or not the flow of the passenger has been disturbed at both the passenger on the conveyer and the entrance / exit of the conveyer, based on the occurrence frequency of each inclination direction value. be able to.

Embodiment 5 FIG. Further, in the above-described embodiment, an example was described in which the movement trajectory was disturbed when the passenger fell, but, for example, the user of the man conveyor (for example, an escalator) erroneously (or intentionally) reverses the driving direction of the man conveyor. When the vehicle enters the direction, a moving trajectory having a completely opposite inclination to the normal band region 33 is obtained as in the band region 39 shown in FIG. Therefore, this abnormal situation can be detected by the frequency evaluation means 9 and the abnormality determination means 10. Also,
For example, it is needless to say that the congestion degree of the passenger can be calculated based on the magnitude of the frequency of occurrence of the inclination direction value of the moving direction-time axis image of the first area 13.

[0050]

According to the first aspect of the present invention, an image pickup means for picking up an image of a passenger on a conveyer to obtain image data, and detecting a disturbance in a moving direction of the image data determined by a moving direction of the conveyer, Since the apparatus is provided with the abnormality determining means for determining an abnormality in the flow of the passengers, it is possible to predict in advance the falling accident of the passengers and to prevent the occurrence of a large accident.

According to the second aspect of the present invention, an image pickup means for picking up an image of a passenger on a conveyor and obtaining image data, and extracting a characteristic image such as a contour of the passenger from the image data,
Flow observing means for arranging the moving direction of the feature image for each transition of time, observing the flow of the passenger accompanying the movement of the conveyor, and turbulence for extracting turbulence from the observed flow of the passenger Because it was configured with extraction means,
By extracting the turbulence of the flow, the occurrence of the accident can be grasped more accurately, so that there is an effect that an appropriate countermeasure for the accident can be taken.

According to the third aspect of the present invention, an image pickup means for picking up an image of a passenger on a conveyer and obtaining image data, and extracting a characteristic image such as a contour of the passenger from the image data,
Arrange the moving direction of the feature image for each transition of time, flow observing means for observing the flow of the passenger accompanying the movement of the conveyor, and inclining the direction of the flow of the passenger observed by the flow observing means. A flow direction extracting unit that extracts the direction value, a tilt direction value extracted by the flow direction extracting unit, and a tilt direction value when the flow of the passenger is steady; Flow direction comparing means for determining the turbulence of the flow of the passenger based on the frequency of extraction of the value, so that the occurrence frequency of the inclination direction value is calculated so that an abnormal situation occurring on the There is an effect that the accident can be grasped and the accident can be minimized.

According to the fourth aspect of the present invention, an image pickup means for picking up an image of the passengers on the man conveyor and the passengers near the entrance / exit of the man conveyor to obtain image data of each passenger, and a contour of each passenger based on the image data. After extracting the characteristic images, the moving directions of the characteristic images are arranged for each transition of time, and the first and second flow observing means for observing the flow of each passenger accompanying the movement of the man conveyor, and First and second flow direction extracting means for extracting the direction of the flow of the passenger on the conveyer and the flow of the passenger near the conveyer entrance and exit, respectively, which are observed by the first and second flow observing means, as inclination direction values; The inclination direction value extracted by both the first and second flow direction extraction means is compared with the inclination direction value when the passenger flow is steady, and based on the frequency of extraction of the inclination direction value other than the steady case. Disturbance of the passenger flow Since it is configured and a flow direction comparing means for judging, passengers passenger conveyor in a wide field of view to predict in advance that lead to abnormal situation, there is an effect that that lead to serious accidents can be prevented as much as possible.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the configuration of a passenger abnormality detection apparatus for a man conveyor according to the invention of claim 1;

FIG. 2 is a block diagram showing one embodiment of a configuration of a passenger abnormality detecting apparatus for a man conveyor according to the invention of claim 2;

FIG. 3 is a block diagram showing one embodiment of the configuration of a passenger abnormality detection apparatus for a man conveyor according to the invention of claim 3;

FIG. 4 is a block diagram showing an embodiment of a configuration of a passenger abnormality detection apparatus for a man conveyor according to the invention of claim 4;

FIG. 5 is a diagram showing each area set by each area setting means 2, 3 according to the present embodiment.

FIG. 6 is a diagram showing a moving direction-time axis image of the passenger generated by the moving direction-time axis image generating means 6 according to the embodiment.

FIG. 7 is a diagram showing a moving trajectory of a passenger generated from a moving direction-time axis image of the passenger according to the embodiment.

FIG. 8 is a diagram illustrating an operation of calculating a tilt direction by a tilt direction frequency calculation unit according to the embodiment.

FIG. 9 is a diagram illustrating an operation of calculating a tilt direction frequency by a tilt direction frequency calculation unit according to the embodiment.

FIG. 10 is a diagram showing a moving direction-time axis image showing a state of a reverse approach operation of the passenger generated by the moving direction-time axis image generating means according to the embodiment.

FIG. 11 is a mechanical diagram showing a mechanical configuration of a conventional passenger abnormality detection device for a man conveyor.

FIG. 12 is a sequence circuit diagram showing an operation control circuit of the conventional device.

[Explanation of symbols]

1 ITV camera 1A flow observation means 1B disturbance extraction means 1C flow direction extracting means 1D flow direction comparing unit 1A ₁ first stream observation means 1A ₂ second flow observation unit 1C ₁ first flow direction extractor 1C ₂ second flow direction extraction Means 17 Passenger 10 Abnormality judgment means

Claims (4)

(57) [Claims] 1. An imaging means for imaging passengers on a man conveyor to obtain image data, and an abnormality for detecting a disturbance in the moving direction of the image data determined by the moving direction of the man conveyor and judging an abnormality in the flow of the passengers. A passenger abnormality detection device for a man conveyor, comprising: a determination unit. 2. An image pickup means for picking up an image of a passenger on a conveyer to obtain image data, and extracting a characteristic image such as a contour of the passenger from the image data, and then changing a moving direction of the characteristic image at each time transition. A conveyor, comprising: flow observing means for arranging and observing the flow of the passenger accompanying the movement of the conveyer; and turbulence extracting means for extracting turbulence of the flow from the observed flow of the passenger. Passenger abnormality detection device. 3. An image pickup means for picking up an image of a passenger on a conveyer to obtain image data, and extracting a characteristic image such as a contour of the passenger from the image data, and then changing a moving direction of the characteristic image at each time transition. Arrange, flow observation means for observing the flow of the passenger accompanying the movement of the conveyor, and flow direction extraction means for extracting the direction of the flow of the passenger observed by the flow observation means as a tilt direction value, The inclination direction value extracted by the flow direction extraction means is compared with the inclination direction value when the passenger flow is steady, and the turbulence of the passenger flow is determined based on the appearance frequency of the inclination direction value other than the steady case. And a flow direction comparing means for judging a passenger abnormality of the passenger conveyor. 4. An image pickup means for picking up an image of passengers on the conveyor and passengers near the entrance / exit of the conveyor and obtaining image data of each passenger, and extracting a characteristic image such as a contour of each passenger from the image data. First and second flow observing means for arranging the moving directions of the feature images for each transition of time and observing the flow of each passenger accompanying the movement of the conveyor.
First and second flow direction extracting means for extracting the direction of the flow of the passenger on the conveyer and the flow of the passenger near the entrance / exit of the conveyer observed by the first and second flow observing means, respectively, as an inclination direction value; The inclination direction value extracted by both the first and second flow direction extraction means is compared with the inclination direction value when the passenger flow is steady, and based on the appearance frequency of the inclination direction value other than the steady case. And a flow direction comparing means for determining the turbulence of the passenger flow.