JP6210946B2 - Angle-of-view adjustment device for camera in elevator car and method of angle-of-view adjustment for camera in elevator car - Google Patents

Description

  The present invention relates to an angle-of-view adjustment device for an in-elevator car camera and an angle-of-view adjustment method for an in-elevator car camera for appropriately setting the depression angle of the camera so that the boarding space in the elevator car is within the field of view.

  It is necessary to appropriately set the depression angle of the elevator car camera so that the boarding space in the car is within the field of view. However, since the size and layout of the car vary from elevator to elevator, it is impossible to uniformly define the depression angle setting standard. Therefore, the adjustment is left to the subjective judgment of the worker.

  On the other hand, in the pan direction (left and right swing) and tilt direction (up and down swing) so that the “model of human appearance” of the image recognition device matches the actual camera image appearance. There is a conventional technique that supports the adjustment (for example, see Patent Document 1).

Japanese Patent No. 5325251

However, the prior art has the following problems.
The conventional technology such as this Patent Document 1 needs to give a lot of data related to the specifications of the elevator and the camera in advance, and there are many operations to be performed in the field, which burdens the operator. Therefore, the possibility of input mistakes and the like is increased, and convenience is lacking.

  In addition, when the camera is operated out of focus, a blurred image is taken, and there is a problem that the function as a crime prevention / recording image is not performed.

  The present invention has been made to solve the above-described problems, and is capable of easily adjusting the camera angle of an elevator car camera by reducing as much data that needs to be given in advance as much as possible. And it aims at obtaining the angle-of-view adjustment method of the camera in an elevator car.

  An angle-of-view adjustment apparatus for an elevator car camera according to the present invention includes an image acquisition unit that acquires an in-car image via a camera installed in the elevator car, and a distortion parameter estimation that estimates a distortion parameter from the in-car image. A distortion correction unit that generates a distortion-corrected image from the in-car image using the distortion parameter estimated by the distortion parameter estimation unit, and a distortion-corrected image generated by the distortion correction unit, The depression angle estimator for estimating the depression angle of the camera, the current depression angle of the camera estimated by the depression angle estimation section, and a target depression angle set in advance are compared, and whether or not the current depression angle of the camera is within the target depression angle range. If the result of the acceptance angle acceptance by the target depression angle comparison unit and the depression angle acceptance determination result by the target depression angle comparison unit fails, a message specifying the direction in which the depression angle should be adjusted A display control unit that displays on the display unit and prompts the operator to adjust the depression angle in an appropriate direction, and the distortion parameter estimation unit acquires a portion that should be a straight line part included in the in-car image as a curve. And a distortion parameter is estimated as a conversion coefficient that makes the extracted portion a straight line without distortion, and the depression angle estimation unit extracts a plurality of straight line portions from the image after distortion correction, The vanishing point position that is the intersection of the straight line parts is calculated, and the depression angle of the camera is estimated based on the preset focal length of the camera and the calculated vanishing point position.

  In addition, the angle-of-view adjustment method for the camera in the elevator car according to the present invention estimates the current camera depression angle from the in-car image acquired via the camera installed in the elevator car, and the estimated current camera depression angle. Is an angle-of-view adjustment method for an elevator car camera for realizing a function of prompting an operator to adjust the depression angle in an appropriate direction when the angle is not within the target depression angle range. An image acquisition step for acquiring an image, a distortion parameter estimation step for estimating a distortion parameter from the in-car image, and a distortion parameter estimated by the distortion parameter estimation step are used to generate a distortion-corrected image from the in-car image. A distortion correction step, a depression angle estimation step for estimating the current camera depression angle from the image after distortion correction generated in the distortion correction step, and The current camera depression angle estimated in the estimation step is compared with the target depression angle set in the storage unit in advance, and the depression angle acceptance / rejection determination is performed based on whether or not the current camera depression angle is within the target depression angle range. When the result of the acceptance / rejection judgment at the comparison step and the target depression comparison step is unsuccessful, a message that defines the direction in which the depression should be adjusted is displayed on the display unit, and the depression adjustment in the appropriate direction is performed by the operator. A display control step for prompting, wherein the distortion parameter estimation step extracts a portion where the portion that should be a straight line portion included in the in-car image is acquired as a curve, and the extracted portion is defined as a straight line without distortion. As a conversion coefficient, the distortion parameter is estimated, and the depression angle estimation step extracts a plurality of straight line portions from the image after distortion correction, and at the intersection of the plurality of straight line portions. It calculates the position of that vanishing point, in which the focal length of the camera set in advance in the storage unit, based on the calculated positions of vanishing points, to estimate the depression angle of the camera.

  According to the present invention, the vanishing point position is calculated from the image after distortion correction, the current camera depression angle is estimated from the known camera focal length and vanishing point position, and the estimated depression angle is within a desired range. If not, it has a function of notifying the operator of the depression angle adjustment direction, thereby reducing the data that needs to be given in advance as much as possible, so that the camera angle adjustment device for the camera in the elevator car can be adjusted easily. And a method of adjusting the angle of view of the camera in the elevator car.

It is a block diagram of the angle-of-view adjustment apparatus of the camera in an elevator car in Embodiment 1 of this invention. It is explanatory drawing regarding the depression angle adjustment of the camera in Embodiment 1 of this invention. It is explanatory drawing regarding the pass / fail determination of a focus by the focus determination part in Embodiment 1 of this invention. It is explanatory drawing of distortion correction in Embodiment 1 of this invention. It is explanatory drawing regarding the vanishing point in Embodiment 1 of this invention. It is a flowchart which shows the angle-of-view adjustment method of the camera in an elevator car in Embodiment 1 of this invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of an angle-of-view adjustment apparatus for an in-elevator car camera and an angle-of-view adjustment method for an in-elevator car camera of the present invention will be described with reference to the drawings.
The present invention can easily perform focus adjustment, image distortion adjustment, and depression angle adjustment based on information on the focal length of a camera to be used, reduces the burden on the operator, and maintains a constant level regardless of the skill of the operator. The above adjustment accuracy can be ensured.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram of an angle-of-view adjustment device for an elevator car camera according to Embodiment 1 of the present invention. The angle-of-view adjustment device for an elevator car camera according to the first embodiment shown in FIG. 1 includes an image acquisition unit 11, a distortion parameter estimation unit 12, a distortion correction unit 13, a data storage unit 14, a depression angle estimation unit 15, and a target depression angle. The comparison unit 16, the focus determination unit 17, and the display control unit 18 are provided.

  Then, the angle-of-view adjustment device for the elevator car camera in the first embodiment performs guidance display on the display unit 2 as necessary based on the image acquired from the camera 1, and then performs focus adjustment and image The technical feature is that distortion adjustment and depression angle adjustment can be easily performed.

FIG. 2 is an explanatory diagram relating to the depression angle adjustment of the camera according to Embodiment 1 of the present invention. First, a normal depression angle setting method will be described with reference to FIG.
(Procedure 1) The depression angle of the camera 1 is turned downward so that the in-car illumination 3 does not enter the field of view.
When the depression angle is small (in the upward direction), the car lighting 3 may appear in the image.

  This is a particular problem in cameras with auto gain control. In a camera with auto gain control, in a situation where the car lighting 3 is reflected in the image, the entire image becomes brighter, so that the gain is lowered and the input image is adjusted to be darker. That is, the range (in-car riding space) that is originally desired to be photographed is dark, and the object (person) to be photographed is dark.

  As a result, the sensitivity of the image recognition apparatus using the car camera is reduced. If the depression angle is small, the area directly under the camera cannot be captured. Therefore, the depression angle of the camera is first adjusted downward.

(Procedure 2) The depression angle of the camera is turned upward so that the in-car indicator 4 installed near the door 5 is in the field of view.
If the depression angle is made too large (too downward) in step 1, the car interior indicator 4 will not be captured, resulting in the area (car interior space) that is originally intended to be taken out of view. Therefore, the depression angle of the camera is adjusted so that the depression angle does not face downward so that the in-car indicator 4 moves away from the field of view.

(Procedure 3) In consideration of the adjustments in Procedures 1 and 2, the depression angle is made as small as possible (upward) so that the car illumination 3 does not enter the field of view.

Based on such a normal depression angle setting method, a concept for quantitatively setting an optimum depression angle according to the present invention will be described.
(1) Target depression angle The target depression angle strictly depends on the following parameters.
-Angle of view of camera 1 (focal length)
・ Installation height of in-car indicator 4 ・ Installation height of camera 1 ・ Ceiling height

  In many cases, the indicator 4 and the camera 1 in the car are both mounted at a high position in the car. Accordingly, if it is assumed that the installation height of the indicator 4 and the camera 1 is the same, the range of depression angles (that is, the target depression angle) in which the indicators are captured can be determined by giving the angle of view.

  Therefore, in the present invention, the depression angle estimation unit 15 estimates the depression angle of the current camera, and the target depression angle comparison unit 16 compares the estimation result with the target depression angle. Then, when the depression angle estimation unit 15 determines that the estimation result is out of the target depression angle range, the display unit 18 causes the display unit to “turn the depression angle upward”, Alternatively, a message such as “Please turn down the corner” is displayed. As a result, the operator can set the actual depression angle within the target depression angle based on the display content.

(2) Determination of Focus The focus determination unit 17 of the present invention determines whether or not the focus is acceptable from the camera image captured via the image acquisition unit 11. For example, the focus determination unit 17 can determine whether or not focus is achieved by determining the threshold of the contrast of the image without using the reference video by the following method.

A specific focus pass / fail determination method by the focus determination unit 17 will be described with reference to FIG. FIG. 3 is an explanatory diagram relating to focus determination by the focus determination unit 17 according to Embodiment 1 of the present invention.
(Procedure 1) As shown in FIG. 3A, the image is scanned in the horizontal direction and the change in luminance is examined.

(Procedure 2) As shown in FIG. 3B, a point at which the luminance gradient is maximized is obtained, and if the gradient at that point is smaller than a preset allowable threshold, the focus is shifted. Is determined.
(Procedure 3) When it is determined that the focus is out of focus, the focus determination unit 17 notifies the display unit 2 via the display control unit 18 that “adjust the focus because it is out of focus”. Is displayed to prompt the operator to adjust the focus.

(3) Estimation of distortion parameters When a linear object is photographed by the camera 1 having lens distortion, the straight line portion appears as a curve due to the influence of distortion. On the other hand, since the in-car camera 1 is generally wide-angle, distortion increases. Therefore, in the present invention, the distortion correction function is realized by the functions of the distortion parameter estimation unit 12 and the distortion correction unit 13.

  In the case of an elevator, it is possible to calculate a distortion parameter such that a curve in an image becomes a straight line by utilizing the fact that the structure in the car is mainly composed of a straight line. Therefore, for example, reference (F. Devernay and O. Faugeras. Straight lines have to be straight Automatic calibration and removal of distortion from scenes of structured environments. Machine Vision and Applications. Vol.13, pp.14-24, 2001. By applying the prior art shown in Fig. 2), the edge of the curved line that should be a straight line extracted from the image is approximated by a polygonal line, and the square error when fitting the straight line to the polygonal line is minimized. Can be obtained from a single image.

  That is, the distortion parameter estimation unit 12 eliminates the need to give parameters in advance by estimating the distortion parameters from the image taken by the camera 1 in the elevator car according to the following procedure. Further, by removing the distortion, there is an advantage that the accuracy of the vanishing point detection process is improved as will be described later.

<Distortion correction procedure>
Assume that an ideal image point (xu, yu) without distortion is moved to (xd, yd) by a distorted lens. In this case, both coordinate values are related by the following equation (1).

  In order to perform distortion correction, it is necessary to estimate K1, cx, and cy in the above equation (1). Therefore, these are collectively referred to as distortion parameters. By estimating the distortion parameter, the distortion can be corrected. Although a more complicated model may be considered, it is known that a simplified model according to the above equation (1) is sufficient in practice.

  FIG. 4 is an explanatory diagram of distortion correction in the first embodiment of the present invention. FIG. 4A shows an image (original image) acquired by the image acquisition unit 11 via the camera 1 and there is distortion. FIG. 4B is an image in which a portion that is originally a straight line portion is extracted as a broken line by a conventional technique as described in a reference document. Further, FIG. 4C is an image after performing distortion correction using the distortion parameter calculated by the above equation (1).

(4) About vanishing point detection When a straight line of a door or a car frame is extracted from a distortion-corrected in-car image, these straight line groups intersect at a single point (below the image). Such a point is called a vanishing point. FIG. 5 is an explanatory diagram relating to the vanishing point in the first embodiment of the present invention. Such a position of the vanishing point in the vertical direction has the following characteristics.

<Features of the upper and lower positions of the vanishing point>
-It does not depend on the distance from the camera 1 to the wall or the door 5.
-It does not depend on the installation height of the camera 1.
Depends on the depression angle of camera 1.
Depends on the focal length of the camera 1

  Therefore, if only the focal length of the camera 1 is given, the depression angle of the camera can be estimated from the position of the vanishing point. Here, the relationship between the position of the vanishing point and the camera depression angle may be given in the form of a table in advance or modeled according to the focal position of the camera.

  Therefore, the depression angle estimation unit 15 calculates the position of the vanishing point from the distortion-corrected image acquired from the distortion correction unit 13. Further, the depression angle estimation unit 15 can obtain the depression angle of the camera from the value regarding the focal length of the camera stored in advance and the position of the calculated vanishing point.

  FIG. 6 is a flowchart showing the angle-of-view adjustment method for the elevator car camera according to Embodiment 1 of the present invention. In summary, the angle-of-view adjustment apparatus for an elevator car camera according to the present invention can easily adjust the angle of view of the camera by executing the following series of procedures according to the flowchart of FIG.

(Step S1) Focus Adjustment Processing The focus determination unit 17 obtains the maximum gradient point from the luminance change of the in-car image acquired from the camera 1 via the image acquisition unit 11, and quantitatively evaluates the focus. FIG. 3 shows an example of obtaining the maximum gradient point from the luminance change in the horizontal direction of the in-car image. When the gradient at the maximum gradient point is smaller than the predetermined threshold value, the focus determination unit 17 displays a message such as “Please adjust because the focus is out of alignment” via the display control unit 18. 2 is displayed to prompt the operator to adjust the focus.

  As a result, the operator can easily adjust the focus of the camera so that the level of the gradient is greater than a predetermined threshold by the adjustment work based on the display content. Note that this focus adjustment process is not an indispensable component, and this function is omitted if focus adjustment has been performed in advance, or if a camera with an autofocus function or a camera with a deep focal depth is used. Is also possible.

(Step S2) Original Image Distortion Correction Processing The distortion parameter estimation unit 12 linearizes the polygonal line approximation of the curved portion from the original image after the focus adjustment by using the conventional technique disclosed in the reference literature. The distortion parameter is estimated using the above equation (1). Further, the distortion correction unit 13 corrects the original image using the distortion parameter calculated by the distortion parameter estimation unit 12, and outputs a corrected image.

(Step S <b> 3) Current depression angle estimation process based on the corrected image The depression angle estimation unit 15 calculates the position of the vanishing point from the distortion-corrected image acquired from the distortion correction unit 13. Further, the depression angle estimation unit 15 estimates the depression angle of the current camera from the value regarding the focal length of the camera stored in the data storage unit 14 in advance and the calculated position of the vanishing point.

(Step S <b> 4) Depression Angle Adjustment Processing The target depression angle comparison unit 16 determines whether or not the current depression angle of the camera 1 estimated by the depression angle estimation unit 15 is within a preset target depression angle range. Then, the target depression angle comparing unit 16 “please make the depression angle upward” with respect to the display unit 2 via the display control unit 18 when the estimation result deviates from the range of the target depression angle, or Displaying a message such as “Please turn the depression angle downward” prompts the operator to adjust the depression angle. As a result, the operator can easily adjust the depression angle of the camera so as to be within the target depression angle range by the adjustment work based on the display content.

  As described above, according to the first embodiment, the operator sequentially performs the focus adjustment process, the distortion correction process, the current depression angle estimation process, and the depression angle adjustment process according to the displayed message as necessary. Thus, appropriate depression angle adjustment can be easily performed. In addition, there is no parameter to be input locally, and by simply setting the focal length in advance, it is possible to achieve depression adjustment satisfying a certain level without individual differences according to the display contents.

  Reference Signs List 1 camera, 2 display unit, 11 image acquisition unit, 12 parameter estimation unit, 13 distortion correction unit, 14 data storage unit, 15 depression angle estimation unit, 16 target depression angle comparison unit, 17 focus determination unit, 18 display control unit

Claims (3)

An image acquisition unit for acquiring an in-car image via a camera installed in the elevator car;
A distortion parameter estimation unit for estimating a distortion parameter from the in-car image;
A distortion correction unit that generates an image after distortion correction from the in-car image using the distortion parameter estimated by the distortion parameter estimation unit;
A depression angle estimation unit that estimates a current camera depression angle from the image after distortion correction generated by the distortion correction unit;
The present camera depression angle estimated by the depression angle estimation unit is compared with a preset target depression angle, and a depression angle acceptance / rejection determination is performed based on whether the current camera depression angle is within the range of the target depression angle. A target depression angle comparison unit;
In a case where the result of the acceptance angle acceptance / rejection determination by the target depression angle comparison unit is unacceptable, a message for defining the direction in which the depression angle should be adjusted is displayed on the display unit, and a display prompting the operator to adjust the depression angle in an appropriate direction A control unit and
The distortion parameter estimation unit extracts a portion where the portion that should originally be a straight line portion included in the in-car image is acquired as a curve, and converts the extracted portion into a straight line without distortion. Estimating the distortion parameters;
The depression angle estimation unit extracts a plurality of straight line portions from the image after distortion correction, calculates a position of a vanishing point that is an intersection of the plurality of straight line portions, and calculates a focal length of the camera set in advance. An angle-of-view adjustment device for an elevator car camera that estimates a depression angle of the camera based on the position of the vanishing point. In the elevator angle-of-view camera adjusting device according to claim 1,
A focus determination unit that determines whether the image is in focus or not based on whether or not the maximum gradient of the luminance change of the in-car image acquired by the image acquisition unit is greater than a predetermined threshold gradient; ,
When the focus determination result by the focus determination unit is unacceptable, the display control unit displays a message notifying that the focus should be adjusted on the display unit, and prompts the operator to adjust the focus.
The distortion parameter estimation unit estimates a distortion parameter from the in-car image after focus adjustment,
The distortion correction unit generates an image after distortion correction from the in-car image after focus adjustment using the distortion parameter estimated by the distortion parameter estimation unit. The current camera depression angle is estimated from the in-car image acquired via the camera installed in the elevator car, and if the estimated current camera depression angle is not within the target depression angle, the depression angle is set in an appropriate direction. A method for adjusting the angle of view of a camera in an elevator car for realizing a function of prompting an operator to perform adjustment,
An image acquisition step of acquiring an in-car image via the camera;
A distortion parameter estimation step for estimating a distortion parameter from the in-car image;
Using the distortion parameter estimated by the distortion parameter estimation step, a distortion correction step of generating an image after distortion correction from the in-car image;
A depression angle estimation step for estimating a current camera depression angle from the image after distortion correction generated in the distortion correction step;
A comparison is made between the current camera depression angle estimated in the depression angle estimation step and a target depression angle set in advance in the storage unit, and whether or not the depression angle is acceptable depending on whether or not the current camera depression angle is within the range of the target depression angle. A target depression angle comparison step for performing the determination;
If the result of the acceptance angle acceptance / rejection determination by the target depression angle comparison step is unsuccessful, a message that defines the direction in which the depression angle should be adjusted is displayed on the display unit, and a display that prompts the operator to adjust the depression angle in an appropriate direction A control step and
The distortion parameter estimation step extracts a portion where the portion that should originally be a straight line portion included in the in-car image is acquired as a curve, and converts the extracted portion into a straight line without distortion. Estimating the distortion parameters;
The depression angle estimating step extracts a plurality of straight line portions from the image after distortion correction, calculates a position of a vanishing point that is an intersection of the plurality of straight line portions, and sets the focal point of the camera previously set in the storage unit An angle-of-view adjustment method for a camera in an elevator car that estimates a depression angle of the camera based on a distance and the calculated position of the vanishing point.

Images