CN103964271B - Elevator monitoring arrangement and elevator monitoring method - Google Patents

Abstract

Provided are an elevator monitoring device and method for processing images of an imaging unit installed obliquely downward in a car to determine the degree of congestion of the car. Use the image captured by the imaging unit installed obliquely downward to take pictures of the inside of the elevator car, and measure the height of a part of the passengers based on the known height mark in the car; based on the image taken by the imaging unit The acquired image tracks the occupant in the car; the position of a part of the occupant in the image on the screen is used as input, and the projection of the part of the occupant when viewed from directly above when projected from the camera unit to the floor of the car is calculated Position: Based on the height of the imaging unit, the height of the above-mentioned part of the occupant, and the positional relationship between the position of the imaging unit and the projection position, calculate the standing position of the occupant when looking down from directly above; estimate the position from directly above based on the standing position of the occupant The occupied area occupied by all passengers when looking down on the inside of the car; and the degree of congestion of the car is determined based on the occupied area.

Description

Elevator monitoring device and elevator monitoring method

technical field

The present invention relates to an elevator monitoring device and an elevator monitoring method.

Background technique

In an elevator car, a state in which there is no room for congestion due to passengers, luggage, etc. is called a full state. In addition, even if it is not full, the congestion state can be quantified according to the width of the congestion room. Furthermore, Patent Document 1 describes that a camera that looks down directly below is installed on the upper portion or side panel of the car to measure changes in floor area occupied by passengers or luggage (paragraph 0006). In addition, Patent Document 2 describes a device that measures the height of a person when the person is riding on the basis of the image captured by the camera in the car looking down directly below, and corrects the height of the person based on the height of the person. The area where the detected passengers match (paragraphs 0043 to 0049). Also, Patent Document 3 describes measuring the height and existing position of a subject based on distance information to an object and camera arrangement information of a stereo camera installed in a car (paragraphs 0039 to 0042). In addition, Patent Document 4 describes a device and method for monitoring a vehicle using a single camera that overlooks a road obliquely below. Here, based on predetermined typical sizes of a large vehicle and a small vehicle, the vehicle position obtained by perspective projection is corrected.

In addition, it is a common method to install a security camera in the elevator car, and it is sought to use the image acquired by the security camera to determine the fullness of the car or measure the degree of congestion. The security cameras are mostly set at the deep corner of the car facing the boarding entrance.

Prior art literature

patent documents

Patent Document 1: Japanese Patent Laid-Open No. 2006-240825

Patent Document 2: Japanese Patent Application Laid-Open No. 8-26611

Patent Document 3: Japanese Patent Laid-Open No. 2001-34883

Patent Document 4: Japanese Patent Laid-Open No. 2008-299458

The problem to be solved by the invention

Using a camera looking down on the floor from the ceiling of the elevator car, as in Patent Document 1, for example, it is relatively easy to measure the floor area occupied by passengers. However, in the image captured by the security camera looking down obliquely from the corner of the car ceiling, in appearance, the floor area occupied by people is measured to be larger than the actual occupied area. Issues such as judging that it is full although it is not full.

In Patent Document 2, in view of the fact that the number of passengers is estimated according to the area of the passengers captured by the camera, or that the closer the distance between the camera and the object is, the area of the passengers becomes larger, by multiplying the correction coefficient corresponding to the height of the passengers, to apply a correction for passenger area. However, the problem of overestimating the degree of congestion depending on the standing position of passengers is also the same as in Patent Document 1, since correction is not performed when looking down from obliquely above.

In Patent Document 3, the height and existing position of an object can be measured based on the stereo distance measurement value. However, in addition to the security camera, it is necessary to add at least two more cameras for stereoscopic distance measurement, so there are problems of an increase in device cost and a reduction in the appearance of the car interior due to the increase of cameras.

Patent Document 4 discloses a device and method for monitoring a vehicle with a single camera overlooking a road obliquely downward. Here, the vehicle position obtained by perspective projection is corrected based on predetermined typical sizes of large vehicles and small vehicles. If this method is directly applied to the elevator car, in the car area, which is an area far narrower than the road, even if a few typical person heights or widths are prepared, the correction accuracy is not sufficient. In addition, unlike vehicles, , there is a problem that it is difficult to judge which character height or width to apply.

Contents of the invention

Therefore, an object of the present invention is to provide an elevator monitoring device and an elevator monitoring method that process images of a camera overlooking obliquely downward, such as a general security camera in an elevator car, and determine the degree of congestion of the car.

means to solve the problem

In order to solve this problem, for example, in an elevator monitoring device provided with an imaging unit installed obliquely downward for imaging the inside of an elevator car, a passenger height measuring unit is provided based on the known inside of the car. The height mark is used as a reference to measure the height of a part of the passenger; the passenger tracking unit tracks the passenger in the car based on the image captured by the imaging unit; the passenger projection unit tracks the passenger The position on the screen of the part of the passenger in the image is used as an input to calculate the position of the part of the passenger viewed from directly above when the imaging unit projects the part of the passenger onto the floor of the car. The projected position at the time; the standing position calculation unit of the passenger, which is based on the height of the imaging unit, the height of the part of the passenger, and the positional relationship between the position of the imaging unit and the projected position , to calculate the standing position of the passenger when looking down from directly above; the occupied area estimating unit, based on the standing position of the passenger, estimates the occupancy of all passengers when looking down on the inside of the car from directly above and a congestion degree determination unit that determines the degree of congestion of the car based on the occupied area.

Invention effect

According to the present invention, the image of the imaging unit installed obliquely downward in the car can be processed to determine the congestion degree of the car. In addition, the problems, configurations, and effects other than those described above will be clarified by the description of the following embodiments.

Description of drawings

Fig. 1 is an example of a configuration diagram of an elevator monitoring device.

Fig. 2 is an example of the flow of processing of the elevator monitoring device.

Fig. 3 shows an installation example of an imaging unit in an elevator monitoring device and an example of an image acquired by the imaging unit.

Fig. 4 is a diagram showing the relationship between the projected position of a person in the elevator monitoring device and the actual standing position.

Fig. 5 is a diagram showing the relationship between the projected position of a person in the elevator monitoring device and the actual standing position in the direction of looking down on the floor of the car directly below.

FIG. 6 is a diagram showing a specific example of projective transformation.

Fig. 7 is a diagram in which the floor occupancy status of the car is estimated by the elevator monitoring device.

detailed description

Embodiments of the present invention will be described with reference to the drawings. In addition, in each figure, the same code|symbol is attached|subjected to the same or similar component, and description is abbreviate|omitted.

Fig. 1 is an example of a configuration diagram of an elevator monitoring device. The elevator monitoring device of the present embodiment is composed of an arithmetic control unit 1 and an imaging unit 2 . The calculation control unit 1 has: a passenger height measurement unit 3 , a passenger tracking unit 4 , a passenger projection unit 5 , a passenger standing position calculation unit 6 , an occupied area estimation unit 7 , a congestion degree determination unit 8 , and an operation control unit 9 , and the output unit 10 . The arithmetic control unit 1 can implement a part or all of the functions by software using an embedded image processing device, a personal computer, or the like. In addition, a part or all of the functions of the arithmetic control unit 1 may be constituted by hardware such as an integrated circuit.

The operation control unit 9 includes a user interface capable of externally controlling the operation control unit 1 to start, stop, or initialize. The output unit 10 has a function of outputting the state of the calculation control unit 1 or the determination result of the congestion degree determination unit 8 to the outside.

Fig. 2 is a diagram showing an example of the flow of processing by the elevator monitoring device. The configuration shown in FIG. 1 will be described based on the flow of processing in FIG. 2 . Immediately after the start of the operation of the elevator monitoring device, each part is initialized (step s01). This is initialization of memory or flags used in the processing of the arithmetic control unit 1 . When realizing the arithmetic control unit 1 with a personal computer, securing and initializing the memory correspond to this. After the initialization, the arithmetic control unit 1 repeats the processing from the image acquisition in step s02 to the congestion degree determination output in step s08. When returning to step s02 from step s08, it is confirmed in step s09 whether there is no end interruption or power off, and if there is no such situation (Yes), the above-mentioned iterative process is carried out, and under the situation of these ( NO) to end the processing.

Before describing the details of the flow of processing in FIG. 2 , an example of installation of the imaging unit 2 and an example of images acquired by the imaging unit 2 will be described based on FIG. 3 . Fig. 3 shows an installation example of an imaging unit in an elevator monitoring device and an example of an image acquired by the imaging unit. Fig. 3(a) is a view of the interior of the elevator car viewed from the direction of the entrance and exit. And, in this example, a security camera is provided as the imaging unit 2 at the upper right corner of the deep part of the car. FIG. 3( b ) and FIG. 3( c ) are examples of images captured by the imaging unit 2 . FIG. 3( b ) shows the moment when the character 20 boards as a passenger when the elevator door is opened, and FIG. 3( c ) shows the situation in which the character 20 boards or stays after the elevator door is closed.

In the process flow of FIG. 2 , when the passenger passes through the entrance and exit of the car, the height h or Height h and width w (step s03). Also as shown in Fig. 3(b), the actual size of the entrance and exit of the car can be measured in advance. Let the height of the entrance and exit be EH and the width of the entrance and exit be EW. In addition, after the imaging unit 2 is installed, it is possible to measure how many pixels each of the height EH and the width EW of the entrance and exit correspond to on the image acquired by the imaging unit 2 . Here, the number of pixels in the height direction of the entrance is Hpix, and the number of pixels in the width direction of the entrance is Wpix.

In step s03 , first, image processing is performed on the image captured by the imaging unit 2 , and the outline or circumscribed rectangle of the person 20 is detected by inter-frame difference and binarization processing of the image captured by the imaging unit 2 . Such image processing itself is a common technique, and thus detailed description thereof will be omitted. Furthermore, the number of pixels for the height and width of the outline or the circumscribed rectangle is obtained respectively. Here, if the number of pixels in the height direction of the person 20 on the image is hpix, and the number of pixels in the width direction of the person 20 is wpix, then it can be calculated based on the ratio of the following formula (1) and formula (2). The height h and width w of the person 20 are measured.

h＝EH×hpix／Hpix...(1)

w=EW×wpix/Wpix...(2)

Immediately thereafter, the person 20 whose height h and the like are measured in step s03 is continuously tracked as a passenger (step s04 ). When new passengers board one after another through the gate, in step 03, the height h etc. are measured by the method described above, and become the object of passenger tracking in step s04. The tracking process is performed by comparing the position on the screen in the previous image with the position on the current screen for each person every time an image is acquired (step s02 ). Therefore, in each loop from step s02 to step s08, the passenger tracking process of step s04 is implemented. On the other hand, the measurement of the passenger's height h in step s03 should be carried out at least once during boarding, so the person whose height h and width w have been measured passes without performing any processing in step s03 . In the passenger tracking in step s04 , the measured height h and the like of the person are followed as attributes of the person. Therefore, for example, even if the person 20 in FIG. 3( c ) stays only for a while in the car, the information on the height h and width w of the person measured during boarding is also retained so that it can be used. In addition, the passenger tracking in step s04 can be based on pattern matching. Note that such tracking processing itself is a common technique, and thus detailed description thereof will be omitted.

In the flow of the process in FIG. 2 , for all the passengers tracked in the car, the projection of the passengers to the floor is carried out by perspective projection (step s05 ) and the standing position of the passengers is calculated (step s06 ).

First, the basic idea will be described using FIG. 4 . Fig. 4 is a diagram showing the relationship between the projected position of a person in the elevator monitoring device and the actual standing position. FIG. 4( a ) is the same as FIG. 3( c ), and is an image of a person 20 staying. In this image, since the tracking is continued from the time of boarding by pattern matching, etc., the top of the head when the height h of the person is measured is found to be at the position shown in A in FIG. 4( a ). For this purpose, the top of the head is projected to the projection position B in FIG. 4( b ) by perspective projection. Furthermore, the distance D from directly below the imaging unit 2 to the projection position B can be calculated. In addition, since the height H from the floor to the imaging unit 2 can be measured after installation, it can be set as a known value if it is measured in advance before the actual processing. Then, after the top of the head A is projected onto the floor, the distance D, the height H, and the height h measured and tracked during boarding are known. The numerical value to be obtained here is the actual standing position, that is, the distance d from directly below the imaging unit 2 to the standing position. Furthermore, the distance d can be calculated based on the proportional calculation of the following formula (3).

d={D×(H—h)}/H...(3)

The projection position on the floor where the top of the head is projected by perspective projection is the same as the projection position of the top of the head in the shadow generated when the point light source is installed at the position of the imaging unit 2 to illuminate the person 20 . Here, a case where the top of the head A of the person 20 at the height h is projected from the imaging unit 2 is taken as an example. As another example, as shown in Fig. 4(c), the height h' of the shoulders can be measured during riding, and the shoulders can be projected onto the floor surface. In this case, the distance D' from directly below the imaging unit 2 to the projected position B' of the shoulder is obtained, and from the distance D', the height H, and the height h', the distance D' from the imaging unit 2 is calculated at the same ratio as before. The distance d' from directly below the part 2 to directly below the shoulder. Strictly speaking, there is a difference between the center of the body directly below the shoulder position and the distance d', but it is almost equal to the distance d (Fig. Calibrate to find the standing position.

In addition, person detection during boarding and pattern acquisition for pattern matching tracking are performed by the passenger height measurement unit 3 . In addition, the pattern tracking of the person is performed by the passenger tracking unit 4 .

Although the above describes the view viewed from the front side (the direction in which the floor is taken as a cross-section), the same conclusion holds true based on the two-dimensional coordinate system looking down on the floor as shown in FIG. 5 . Fig. 5 is a diagram showing the relationship between the projected position of a person in the elevator monitoring device and the actual standing position in a direction overlooking the floor of the car directly below. In FIG. 5 , B is the projection position when the top of the head A of the person 20 indicated by the shaded area is projected from the imaging unit 2 onto the floor surface. The grid area 20' indicates the position where the silhouette of the person 20 is projected. The origin is taken on the floor surface directly below the imaging unit 2, and X-axis and Y-axis as coordinate axes are set along both walls. Assuming that the coordinates of the projection position B are (Dx, Dy) and the standing position of the person 20 is (dx, dy), they have the following relationship similar to the expression (3).

dx={Dx×(H—h)}/H...(4)

dy={Dy×(H—h)}/H...(5)

The height H is known from the beginning, the height h is known when boarding, and (Dx, Dy) is known by perspective projection. Therefore, the standing position (dx, dy) of the person 20 can be calculated by Equation (4) and Equation (5).

Next, the processing in the projection of the occupant to the floor (step s05 ) will be described using FIG. 6 . FIG. 6 is a diagram showing a specific example of projective transformation. FIG. 6( a ) is a diagram showing a bird's-eye view of the car from directly above, and FIG. 6( b ) is a diagram showing an image captured by the imaging unit 2 . How a desired position on the image of FIG. 6( b ), for example, the top of the person's head A is projected onto the floor of FIG. 6( a ) viewed from directly above is a problem, and projective transformation is used to solve this problem.

In FIG. 6(a), the origin O is defined on the floor directly below the imaging unit 2, the X axis is defined parallel to the wall facing the elevator door of the car, and the Y axis is defined parallel to the left wall of the car. Set the unit of length to meters. On the other hand, in Figure 6(b), the upper left of the screen is taken as the origin 0, the horizontal direction of the screen is taken as the x-axis, and the vertical direction is taken as the y-axis. The unit of length is set to number of pixels. Furthermore, the coordinates in FIG. 6( a ) are expressed as (X, Y) ^T , and the coordinates in FIG. 6( b ) are expressed as (x, y) ^T . The T in the upper right corner indicates the transpose matrix. The transformation from (x, y) ^T to (X, Y) ^T is a combination of rotation transformation, translation transformation, and projective transformation, but it can be seen that they are passed through formula (6), or by formula (6) through formula (6) ' and transformed into formula (6) "expression. In addition, formula (6)" is to eliminate the expression of K. Here, (x, y) ^T may also be referred to as camera coordinates (coordinates on an image), and (X, Y) ^T may be referred to as ground coordinates (coordinates when viewed from directly above).

$\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}\\ \mathrm{<span\; class="notranslate">\; K</span>}\end{array}\right)<span\; class="notranslate">\; =</span>\left(\begin{array}{ccc}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}& {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 12</span>}}& {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 13</span>}}\\ {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; one</span>}}& {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; two</span>}}& {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; three</span>}}\\ {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 31</span>}}& {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 32</span>}}& \mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right)\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; Y</span>}\\ \mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right)<span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

$\left.\begin{array}{c}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 13</span>}}\\ \mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; one</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; two</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; three</span>}}\\ \mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 31</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 32</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right\}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>}^{<span\; class="notranslate">\; \&prime;</span>}$

$\left.\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 13</span>}}}{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 31</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 32</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; one</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; two</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; three</span>}}}{{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 31</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 32</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\end{array}\right\}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}$

In formula (6), etc., K is a variable related to the distance between the imaging unit 2 and the subject, and C ₁₁ to C ₃₂ are determined according to so-called camera parameters such as the height, orientation, and magnification of the imaging unit 2 from the floor. Numeric values can be obtained in advance. Here, C ₁₁ to C ₃₂ can obtain parameters C ₁₁ to C ₃₂ by preparing four or more known sets of camera coordinates and ground coordinates by so-called camera calibration. It is required to take more than 4 sets of known coordinate values. For example, mark more than 4 points that are not on a straight line on the floor of the elevator. The coordinates of the marked position are used to obtain the camera coordinates, which can be obtained.

After C ₁₁ ~ C ₃₂ become known through camera calibration, formula (7) obtained by transforming formula (6)" can transform any coordinate (x, y ) ^T is transformed into coordinates (X, Y) ^T when viewed from directly above in FIG. 6( b ).

$\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}\\ \mathrm{<span\; class="notranslate">\; Y</span>}\end{array}\right)<span\; class="notranslate">\; =</span>{\left(\begin{array}{cc}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 31</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; x</span>}& {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 12</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 32</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; x</span>}\\ {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; one</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 31</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}& {\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; two</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 32</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}\end{array}\right)}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\left(\begin{array}{c}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; 13</span>}}\\ \mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; twenty\; three</span>}}\end{array}\right)<span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

Therefore, in the process of projecting the occupant onto the floor (step s05), the occupant projection unit 5 calculates the coordinates of the top of the head A on the image in FIG. Coordinates when viewed from directly above the projection position B when the part 2 projects the top of the head A onto the floor.

Next, in the process of calculating the standing position of the passenger (step s06), the standing position calculation unit 6 of the passenger uses Equation (3) or Equations (4) and (5) based on the height H of the imaging unit 2 and the height of the passenger. The height h of the person, the positional relationship between the position of the imaging unit 2 and the projection position B (such as distance D, coordinates (Dx, Dy), distance Dx, distance Dy, etc.), to calculate The coordinates of the standing position of the passenger when looking down (in this case, the coordinates of the top of the head A). In addition, Equation (3) can be used when using a polar coordinate system, and Equation (4) and (5) can be used when using an XY coordinate system.

The projection of the occupant to the floor (step s05 ) and the calculation of the standing position of the occupant (step s06 ) described above are performed for all occupants in the car.

In addition, as in the example using the shoulders in FIG. The projection unit 5 receives the coordinates of the part A of the occupant on the screen corresponding to the height h, and can obtain the projection position B when viewed from directly above when the part A of the occupant is projected from the imaging unit 2 on the floor. That's it. In addition, some of the passengers to be projected are expected to have as high a portion as possible, such as the top of the head, where it is difficult to hide in a crowded place. Furthermore, the standing position calculation unit 6 of the passenger can calculate the position when viewed from directly above based on the height H of the imaging unit 2, the height h of a part A of the passenger, and the positional relationship between the position of the imaging unit 2 and the projection position B. The coordinates of the standing position of the occupant (in this case, the coordinates of a part A of the occupant, but if it is known to be a specific part such as the shoulder, for example, the coordinates from the center position of the occupant to the part A can be used The coordinates of the central position of the passenger may be corrected to the coordinates of the center position of the passenger and used as the coordinates of the standing position of the passenger).

Next, the estimation of the occupancy area of the passengers and the determination of the degree of congestion in the flow of the process of FIG. 2 will be described (step s07 ). Fig. 7 is a diagram showing the floor occupancy status of the car estimated by the elevator monitoring device. FIG. 7( a ) is a conceptual diagram of an image obtained when the ceiling of the car is assumed to look down from a relatively high position. The rectangular area represents the car area 30, and the stippled area is an area 31 where the floor of the car can be seen. Furthermore, the shaded portion is the region 32 where the passenger occupies the floor. It can be considered that as the proportion of the area 31 decreases, the degree of congestion increases. Furthermore, if the total area of the area 31 is less than one person, it can be considered that the area is congested to the extent that there is no room, and it can be judged to be full. According to this embodiment, like a general security camera, the image obtained by the imaging unit 2 installed at the corner of the car can be processed to obtain an image like that shown in FIG. Based on the occupancy rate of the floor surface, the congestion degree of the car is judged.

FIG. 7( b ) is an image in which the occupancy status of the floor of the car is estimated based on the standing position of the passenger obtained in step s06 . The circle area 33 of the lattice pattern centering on the standing position of the passenger (the coordinate of A) and having the width w of the passenger as the diameter is an area estimated to be occupied by a person. Furthermore, the area 31 is an area where the floor can be seen, that is, an area where there is room for congestion. The degree of congestion is determined based on the size of the area, and it is determined that the area 31 is full when the area 31 is less than one person. Here, as the width w, the width measured during boarding may be used, or the measurement of the width w during boarding may be omitted, and a predetermined predetermined width set in advance may be used.

In fact, most of the occupancy area occupied by passengers is not a circle, but an elliptical area with the two shoulders as the long axis. For this reason, the ratio of the long axis to the short axis when the occupancy area of the person is approximated by an ellipse can be calculated statistically in advance, and the occupancy ellipse with the width w as the long axis can be arranged at the standing position, Estimation is performed as shown in FIG. 7( c1 ) and FIG. 7( c2 ). FIG. 7( c1 ) aligns the orientation of the passenger's face with the short-axis direction of the elliptical region 33 . In order to realize this, for example, when the passenger is tracked in step s04 , it is sufficient to detect the facial features and detect the orientation of the face. On the other hand, in FIG. 7( c2 ), the area 33 of the ellipse is arranged irrespective of the direction of the face of the passenger. In this way there is no need to detect the orientation of the face. In addition, even if the passenger's orientation is different from the actual one, there is no big difference in the area occupied by the floor surface of the car. Therefore, to determine the degree of congestion or full occupancy of the car, the elliptical area 33 is arranged regardless of the passenger's orientation. enough.

For example, as shown in FIG. 7 , the occupied area estimating unit 7 creates an estimate representing the area occupied by the passenger when looking down on the inside of the car from directly above, based on the standing position of the passenger obtained by the standing position calculating unit 6 of the passenger. image, and estimate the occupancy area occupied by all occupants.

The congestion degree determination unit 8 determines the congestion degree based on the occupied area of all passengers estimated by the occupied area estimation unit 7 . As an example of the degree of congestion, for example, it can be expressed by the ratio of the occupied area of all passengers estimated by the occupied area estimating unit 7 to the area of the car known in advance, or the ratio or the occupied area It divides into several stages and expresses the degree of congestion in a plurality of levels such as large, medium, and small, or 1, 2, and 3. In addition, if the case where the degree of congestion is the largest is considered to be full, the determination of full capacity may be included in the determination of the degree of congestion. In addition, it is also included in the concept of judging the degree of congestion that only judges whether it is full or not. As a method of judging full occupancy, there are cases where the area of the floor (subtracting the occupancy area of passengers from the area of the car) is smaller than a given threshold (for example, per person), or if the degree of congestion or the occupancy area is equal to that indicating full occupancy The method of judging that it is full if it exceeds a given threshold, etc.

Next, in the process of the congestion degree judgment output (step s08 ), the output unit 10 outputs the judgment result of the congestion degree judgment unit 8 (the degree of congestion, whether it is full or not, etc.).

Then, the elevator control device (not shown) controls the elevator based on the congestion degree determination output. For example, even if a hall call is registered on a floor that is not scheduled to stop when it is full, control such as passing without responding to a hall call is performed.

As mentioned above, according to this embodiment, the image of the imaging part installed diagonally downward in a car is processed, and the degree of congestion of a car can be judged.

In addition, this invention is not limited to the above-mentioned embodiment, Various modification examples are included. For example, the above-mentioned embodiment has been described in detail for the purpose of explaining the present invention intelligibly, but it is not necessarily limited to having all the configurations described above.

Symbol Description

1···Operation control unit

2···camera department

3···Passenger Height Measurement Department

4··· Passenger Tracking Department

5··· Passenger projection unit

6···Standing position calculation unit of passengers

7···Occupied Area Estimation Department

8···Congestion Judgment Unit

9···Operation Control Department

10···Output

20···Character

Claims (6)

1. an elevator monitoring arrangement, possesses in the car to elevator the image pickup part arranged obliquely downward taken,

The feature of described elevator monitoring arrangement is to have:

Passenger height meter, it measures the height of a part for passenger for benchmark with the mark of the known height in described car;

Passenger tracking portion, it follows the tracks of described passenger based on the image photographed by described image pickup part in described car;

Passenger Projection Division, its using the position of a described part for the described passenger in described image on picture as input, calculate when a described part for described passenger being projected to the floor of described car from described image pickup part from directly over observe time projected position;

The standing place calculating part of passenger, the height of a described part of its height based on described image pickup part, described passenger and the position relationship between the position of described image pickup part and described projected position, calculate from directly over overlook time the standing place of described passenger;

Occupied area estimator, it is based on the standing place of described passenger, estimate from directly over overlook in described car time whole passenger shared by occupied area; And

Degree of congestion detection unit, it judges the degree of congestion of described car based on described occupied area.

2. elevator monitoring arrangement according to claim 1, is characterized in that,

Described passenger height meter also measures the width of described passenger,

Described occupied area estimator also utilizes the width of described passenger to estimate the occupied area shared by described whole passenger.

3. elevator monitoring arrangement according to claim 1 and 2, is characterized in that,

Whether described degree of congestion detection unit judges in described car full.

4. an elevator monitoring method, uses in the car to elevator the image taken by the image pickup part arranged obliquely downward taken to monitor,

The feature of described elevator monitoring method is,

Measure the height of a part for passenger for benchmark with the mark of the known height in described car; Based on the image photographed by described image pickup part, in described car, follow the tracks of described passenger; Using the position of a described part for the described passenger in described image on picture as input, calculate when a described part for described passenger being projected to the floor of described car from described image pickup part from directly over observe time projected position; Based on the height of described image pickup part, the height of a described part for described passenger and the position relationship between the position of described image pickup part and described projected position, calculate from directly over overlook time the standing place of described passenger; Based on the standing place of described passenger, estimate from directly over overlook in described car time whole passenger shared by occupied area; And based on described occupied area, judge the degree of congestion of described car.

5. elevator monitoring method according to claim 4, is characterized in that,

Also measure the width of described passenger, also utilize the width of described passenger to estimate the occupied area shared by described whole passenger.

6. the elevator monitoring method according to claim 4 or 5, is characterized in that,

Whether full the judgement of described degree of congestion to comprise in described car judgement.