JP4189755B2 - Method and apparatus for adjusting brightness of camera image - Google Patents

Description

  The present invention relates to a brightness adjustment method for a camera image that can capture a characteristic part (such as a human face) of an object with an appropriate exposure amount when an object such as a person enters the monitoring area of the camera. And the apparatus.

When monitoring whether an object has entered a certain area, a camera is installed, the area is constantly photographed, and the photographed image is processed to determine whether the object has entered.
JP-A-9-319879

However, in the area where the camera is installed, the condition of the light beam may be different for each part.
When an object passes through that area, if shooting is performed following the movement of the object, if the brightness of the object changes suddenly, the responsiveness of the exposure conditions of the camera cannot follow this, and the object is properly captured. There is a problem that you can not.

For example, when an object suddenly appears from a dark place to a bright place, the object appears whitish for a while, and the characteristics of the object cannot be determined. The same applies when an object suddenly enters a dark place from a bright place.
Therefore, the present invention predicts the condition of the light beam in the next moving region even when the object moves in the region where the light beam condition is different for each part, and the image of the object is always set to an appropriate brightness. It is an object of the present invention to provide a brightness adjustment method and apparatus for camera images that can be processed.

According to the brightness adjustment method of the camera image of the present invention, the screen of the monitoring area of the camera is divided into blocks according to the intensity of the light beam, the monitoring area is photographed by the camera, and the light amount is determined for each block. the detecting an object entering the surveillance area, when the object intrusion is detected, in the screen of the camera to track the movement of the set area to contain the object, to the monitoring area When an intruding object moves between blocks in the camera screen, the light beam is weak so that the stronger the light beam, the lower the screen brightness, depending on the light intensity of the block to which the area moves. This is a method of adjusting the luminance so that the luminance of the screen increases as the block increases.

The object may be a person, and movement of an area set to include a human face may be determined within a camera screen.
According to this method, the image of the monitoring area of the camera is divided into blocks according to the intensity of the light beam, and the brightness of the screen is adjusted according to the intensity of the light beam of the block to which the object will move. Can do. This brightness adjustment method can be said to be feed-forward brightness adjustment in the sense that the brightness control area predicts in advance the intensity of the light beam of the block to be moved.

Therefore, when an object moves in a region where the light ray conditions differ from one part to another, the light ray condition of the next moving region can be predicted, and image processing is performed so that the image of the object always has appropriate brightness. be able to.
When the backlight is determined according to the positional relationship between the intruding object, the camera, and the direction of the light beam, the luminance may be adjusted according to the backlight.

The brightness adjustment apparatus for a camera image according to the present invention divides the image of the monitoring area and the image of the monitoring area of the camera into blocks according to the intensity of the light beam, and determines the light amount for each block. comprising a block division unit, a detecting means for detecting an object entering into the monitoring area, when the object intrusion into the surveillance region is detected by the detecting means, in the screen of the camera, the object In accordance with the intensity of the light beam of the block to which the area moves , when the intruding object moves between the blocks in the camera screen, the movement area prediction means for tracking the movement of the area set as described above , Luminance adjustment means for adjusting the luminance is provided so that the luminance of the screen is lowered as the block with the stronger light beam is lower, and the luminance of the screen is higher as the block with the weaker light beam.

The object may be a person, and the brightness adjusting means may determine movement of an area set to include a human face within a camera screen.
The brightness adjusting unit may adjust the brightness according to the backlight when the backlight is determined according to the positional relationship between the camera, the direction of the light beam, and the object.
This camera image brightness adjusting apparatus is an apparatus according to the same invention as the camera brightness adjusting method, and has the same effects.

Hereinafter, an embodiment of the present invention when a house entrance is taken as an example of a monitoring area will be described in detail with reference to the accompanying drawings.
FIG. 1 is a perspective view of the entrance 1 and is also a view for defining a coordinate system. The x and y directions are taken horizontally and the z direction is taken vertically. The door 2 is on the surface of y = 0, and a clog box 3 is installed near the wall of x = 0. Above the clog box 3 is a large window 4 for taking in light rays.

  FIG. 2 is a system configuration diagram of the security monitoring device, which is composed of a security camera 11, a computer 12, an alarm buzzer 13, a mobile phone 14 installed in the room, and an Internet line connected to the outside. The The security camera 11 monitors the entrance 1 by taking pictures periodically. When the computer 12 detects a human intrusion, the computer 12 sends an alarm signal to the alarm buzzer 13 or sends an alarm message to the mobile phone 14.

FIG. 3 is a block diagram illustrating the brightness adjustment function of the computer 12 provided in the crime prevention monitoring apparatus.
The computer 12 includes a video monitoring function B1, a background creation / template creation function B2, an intruder determination function B3, a moving area prediction function B4, a screen light map creation function B5, and a brightness control function B6.

In the video monitoring function B1, the computer constantly monitors images taken by the security camera 11.
In the background creation / template creation function B2, the computer performs a background creation process and a template creation process for the monitoring image. This background creation process is a process of storing an image of the entrance in the absence of a person as a background image (Norio Saeki, Kenshi Suzuki, Kiyoshi Iwata “Intruder Detection Method by Image Recognition” [online], 1999 [Search on April 26, 2004], Internet <http://www.fujitsugeneral.co.jp/japanese/techj/1999-01/pdf/05rgazou.pdf>). The template creation process is a process for preparing and preparing a plurality of templates having a predetermined size in anticipation of a person appearing.

  The screen ray map creation function B5 is a function of dividing the background image into several blocks according to the intensity of the rays. The intensity of this light beam cannot be directly measured by the luminance distribution that appears in the background image. This is because the light beam itself cannot be captured by the camera. Therefore, the angle of the light beam inserted from the window is predicted, and the portion where the light beam passes and the portion where the light beam does not pass are divided into blocks. Since the angle of the light beam inserted from the window is determined by the season, date, and time, the angle of the light beam can be predicted based on the time indicated by the clock incorporated in the security monitoring device. Further, the angle of the light beam inserted from the window may be measured, and this measuring method will be described later with reference to FIG.

Then, the light intensity is estimated for each block. For this purpose, it is advisable to pay attention to an object that is exposed to light, for example, a wall or a floor of an entrance, and monitor the luminance of that part to indirectly estimate the light intensity. Thereby, the intensity of the light according to the difference in the amount of solar radiation such as clear, cloudy, rain, etc. can be accurately measured.
FIG. 4 is a diagram showing an image divided according to the intensity of light in this way. A block drawn with cross-hatching corresponds to a weak portion of light rays, and a block without hatching corresponds to a strong portion of light rays.

In the intruder determination function B3 in FIG. 3, the computer determines an intruding object. This determination function will be described in detail. A background separation process is performed to subtract the background image from the entrance image, and it is detected whether a new object image has appeared in the area of the door 2 of the entrance image taken by the security camera 11.
When a new object image appears, the computer applies the template to determine whether the object is a person.

After detecting a person in this way, the detected image is registered and used as a template.
Using this template, a portion (a brightness control area) for which a clearest image is desired to be obtained, for example, a face portion at the top of the template, is captured, and the moving speed of the portion is measured and tracked. FIG. 4 shows the brightness control areas S0, S1, and S2 of the template. As people walk, they move from S0 to S1 to S2.

The movement area prediction function B4 is a function for predicting which block of the ray map the luminance control area being tracked is about to move to. Taking FIG. 4 as an example, out of the luminance control areas S0, S1, and S2, S0 is about to move from a light-block to a light-block. S1 and S2 are moving in a block with strong light rays.
The brightness control function B6 is a function for controlling the brightness of the entire screen according to the intensity of the light beam of the block whose brightness control area is about to move by the movement area prediction function B4.

For example, the block shown by cross hatching in FIG. 4 is a weak portion of light rays, and when a luminance control area appears there, it appears as low luminance, so that the luminance of the entire screen is increased to make a bright screen. A block without hatching is a portion where light rays are strong, and when a luminance control area appears there, it is reflected in high luminance, so that the luminance of the entire screen is increased to make a dark screen.
The backlight determination function B7 determines the backlight based on the positional relationship among the intruding object, the camera, and the direction of the light beam. Details of the function will be described later.

Hereinafter, the flow of entrance detection and brightness control processing will be described with reference to a flowchart (FIG. 5).
The computer initially sets the brightness control area at the entrance of the entrance, and adjusts the brightness of the screen so that the portion is most clearly seen (step S1). The entrance image taken by the security camera 11 is input (step S2), and the background image creation / template creation processing described above is performed (step S3).

  It is determined whether or not a difference between the entrance image captured by the security camera 11 and the background image has appeared (step S4). If this difference appears, whether or not the size is about the size of a person by template matching. Determine (step S5). If it is about the size of a person, the person's face is tracked (step S6). If tracking is not possible, the process returns to the image input in step S2.

If the tracking is successful (YES in step S7), it is determined whether or not the human face portion is about to move to a different block of rays (step S8).
If the light beam moves to a different block, the brightness of the entire screen is adjusted before the movement in accordance with the light amount of the block area after the movement (step S9).
A method for adjusting the brightness of the entire screen will be described. FIG. 6 is a block diagram showing a brightness adjustment method. When the brightness control area moves to the next block (C2 in FIG. 6), the aperture control correction value of the security camera 11 is set in accordance with the intensity of the light beam of the block. Feed forward control is performed (C3). An aperture control value is determined by this feedforward control (C4). Here, the “aperture control value” refers to any one of shutter speed, exposure diameter, gamma correction value, or a combination thereof.

Further, conventional feedback is combined with this feedforward control. That is, the luminance average value of the current luminance control area is detected (C5) and compared with the target luminance (C6). The result of this comparison is added to the aperture control value (C4) as an aperture control value (feedback control value) based on a luminance difference.
With the above brightness adjustment, for example, when moving from the area near the door 2 where the person enters to the area where the light beam is inserted from the window 4, the face of the person becomes bright and white blurring occurs. By reducing the overall brightness just before the movement, it is possible to clearly see the brightness of a person's face. Conversely, when a person moves from the area where the light is inserted to the dark area behind the entrance, the person's face suddenly becomes dark, so by increasing the brightness of the entire screen just before moving, Can be clearly seen.

  Table 1 summarizes the processing for raising and lowering the brightness of the entire screen depending on whether the brightness control area moves to a block with a strong light beam or a block with a weak light beam.

With the above processing, when a person enters the house through the door 2, the person's face can always be photographed with appropriate brightness regardless of the state of light in the room.
Next, in the case where the intrusion of the object into the monitoring area is detected, when the backlight is determined according to the positional relationship between the direction of the camera, the light beam, and the object, the brightness adjustment according to the backlight is performed. explain.

FIG. 7 is an explanatory diagram for explaining a method of calculating the direction of light rays. (A) is the figure image | photographed from the camera, (b) is the projection figure in a vertical surface.
When calculating the direction of the light ray, it is determined not in the screen of the security camera 11 but in real space (three-dimensional coordinates in the room). For this purpose, it is necessary to grasp in advance the relationship between the position of the security camera 11 to be photographed on the screen and the position of the real space. For this purpose, it is necessary to incorporate a conversion formula between the two-dimensional coordinates on the screen of the security camera 11 and the actual three-dimensional coordinates into the program of the computer 2 using the position and posture of the security camera 11 as parameters ( (See JP-A-4-20189, JP-A-6-35443, etc.). However, in order to convert two-dimensional coordinates on the screen into three-dimensional coordinates, it is necessary to reduce the number of coordinates in the three-dimensional space by one. For this purpose, an object whose coordinates in the three-dimensional space are known, for example, the door 2, the shoebox 3, the window 4, etc. may be used as the reference position.

As a result of the coordinate conversion, the height of the window 4 is L, and the height of the light reflected on the wall is r. The direction θ when the light ray is projected in the vertical plane is
θ = tan ⁻¹ [(L−r) / w]
It can be calculated with W is the width of the entrance.
The height of a person's face is set to h, and the boundary between a block with a strong light beam and a block with a weak light beam can be defined by whether the person's face enters the light beam inserted through the window 4. In FIG. 7B, in a region where the distance from the wall to the person is a and (r + atan θ) ≦ h is satisfied, no light directly hits the face, so the boundary between the strong light block and the weak light block is From the wall facing the window 4, it can be seen that the line is a distance a = W1. The height h of the person's face is determined by determining the height of the intruder based on a comparison between the height of the intruder when the door 2 is opened and the height of the door 2 (known). The face height h can be determined by multiplying the ratio (statistical ratio between face height and height).

The direction φ (see FIG. 8B) when the light ray is projected in the horizontal plane is
φ = tan ^-1 [v / w]
It can be calculated with v is the distance when the boundary of the light reflected on the wall is measured from the window 4 in the y direction.
FIG. 8 is an explanatory diagram for determining backlight. (A) is the figure image | photographed from the camera, (b) is the projection figure in a horizontal surface.

The optical axis of the camera is assumed to be in the −y direction. The difference (π / 2−φ) between the optical axis direction of the camera and the direction φ when the light beam is projected in the horizontal plane is calculated. If this angle is less than 90 °, it is determined that the light is back.
Under the condition of backlight, the brightness control area does not become a whitened image when the brightness control area enters a block with strong light rays (a portion not shaded in FIG. 8B), as indicated by reference numeral S3. As described above, the control is performed by calculating the optimum aperture control value in the direction of darkening the image from the brightness of the wall or floor on which the light is projected and the current aperture control value. When the brightness control area enters a block with weak light rays (the shaded portion in FIG. 8B) as indicated by reference numeral S4, an intruder can prevent the brightness control area from becoming a blackout image. Control is performed so as to obtain the aperture control value when the entrance is monitored just before entering.

  Although the embodiments of the present invention have been described above, the embodiments of the present invention are not limited to the above-described embodiments, and various modifications can be made within the scope of the present invention.

It is the perspective view which looked at the entrance which is a monitoring area. It is a system block diagram of a crime prevention monitoring apparatus. It is a block diagram which shows the function of the computer with which the crime prevention monitoring apparatus was equipped. It is the entrance image figure divided into blocks according to the strength of light. It is a flowchart for demonstrating the flow of the intrusion detection of a front door, and the whole brightness | luminance control processing. It is a block diagram which shows the luminance adjustment method. It is a figure for demonstrating the method to calculate the direction of a light ray. (A) is the figure image | photographed from the camera, (b) is the projection figure in a vertical surface. It is explanatory drawing for backlight determination. (A) is the figure image | photographed from the camera, (b) is the projection figure in a horizontal surface.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Entrance 2 Door 3 Clog box 4 Window 11 Security camera 12 Computer S0, S1, S2, S3, S4 Brightness control area

Claims (4)

The screen of the camera monitoring area is divided into blocks according to the intensity of the light beam,
Shoot the monitoring area with a camera,
Determine the amount of light for each block,
Detecting the intrusion of an object into the monitoring area,
When an intrusion of an object is detected , the movement of the area set to include the object is tracked within the camera screen,
When an object that has entered the monitoring area moves between blocks in the camera screen, the brightness of the screen is lower for blocks with stronger rays depending on the intensity of the rays of the block to which the area moves. A method for adjusting the brightness of a camera image, wherein the brightness is adjusted so that the brightness of the screen increases as the light beam blocks become weaker. A camera that captures the surveillance area;
Block dividing means for dividing the image of the monitoring area of the camera in advance according to the intensity of the light beam and determining the light amount for each block ;
Detection means for detecting an object entering into the monitoring area,
If the object entry into the monitoring region is detected by the detecting means, in the screen of the camera, a moving region prediction means for tracking the movement of the set area to contain the object,
When an intruding object moves between blocks in the camera screen, the light beam is weak so that the stronger the light beam, the lower the screen brightness, depending on the light intensity of the block to which the area moves. A brightness adjustment apparatus for a camera image, comprising brightness adjustment means for adjusting brightness so that the brightness of the screen increases as the block increases. The object is a person,
3. The camera image brightness adjusting apparatus according to claim 2, wherein the moving area predicting means determines movement of an area set to include a human face in the camera screen. Said luminance adjusting means, a camera and when it detects backlight in accordance with the positional relationship between the direction and the object beam, the brightness adjusting device according to claim 2 or claim 3, wherein the camera image to the brightness adjustment corresponding to the backlight .

Images