JP2001333420A - Image supervisory method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image supervisory device that monitors a dark scene with an extended dynamic range and a reduced lighting power. SOLUTION: An image input control section 500 sets a 1st image with normal exposure at strong lighting and a 2nd image with high-speed exposure at weakend lighting to a supervisory scene photographed by an ITV camera 100 with an LED lighting device 200, and an image input section 1000 receives the 1st image and then receives a 2nd image at a succeeding field. An image synthesis section 2000 sums the 1st and 2nd images with a weight coefficient corresponding to the luminance by each pixel multiplied therewith to generate a composite image thereby obtaining a processing object image for object discrimination. A differential image generating section 3000 conducts differential processing by each pixel between a just preceding frame and a current frame of the composite image, and a change area extract section 4000 applies binary processing to the differential image to extract a change area. An object detection section 5000 conducts contrast pattern matching for each inter-frame normalization processing per integrated change area to discriminate and object.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention
By capturing a normal exposure image with the illuminance of the D illuminator increased and a high-speed exposure image with the illuminance of the LED illuminator weakened, and combining them into an image to be processed, continuous monitoring day and night with low current consumption is achieved. The present invention relates to an image monitoring device that performs the monitoring.

[0002]

2. Description of the Related Art There is known an imaging apparatus for obtaining an image of a moving object having a wide dynamic range (Japanese Patent Laid-Open No. 7-75026). Here, the first image captured at the first exposure time and the second image
In the second image captured at the exposure time of, two types of weighting functions (f and g) corresponding to the luminance value of the second image are generated. Then, the luminance value (for each pixel) of the first image is weighted by the function f, and the luminance value (for each pixel) of the second image is weighted by the function g. The composite image created by calculating the luminance value is set as the processing target image.

[0003]

As described above, the first image captured at the first exposure time and the second image captured at the second exposure time are respectively weighted differently. The device for realizing the synthesis
This is effective when the illuminance equal to or more than a predetermined value is guaranteed. However, in a scene where light does not reach during the night or day, it is necessary to turn on the lighting device, and there is no consideration that it is necessary to control turning on and off the lighting device and to control the intensity of the illuminance. For this reason, in continuous monitoring during the day and night, there is a problem that the lighting cost is increased and the life of the lighting device is shortened.

[0004] An object of the present invention is to overcome the above-mentioned problems of the prior art, input an image with an increased illuminance and an image with a reduced illuminance, and combine them to obtain an image to be processed.
It is an object of the present invention to provide an image monitoring device capable of continuous monitoring day and night with low current consumption.

[0005]

According to the present invention to achieve the above object, at the time of image input, a first image in which the illuminance of the illumination device is increased and a second image in which the illuminance of the illumination device is reduced or turned off are captured. Thus, there is a feature in that these are combined into an image to be processed.

In the image surveillance method of the present invention, first, when photographing a surveillance scene with an ITV camera or the like with a lighting device, the first
After inputting the normal exposure image by increasing the illuminance of the illuminating device as the second image, capturing the high-speed exposure image shifted by one field by decreasing the illuminance of the illuminating device as the second image, and obtaining the first image and the second image. Are added with a weighting factor corresponding to the luminance value of each pixel to create a composite image. Here, the illumination of the first image is an illuminance that allows an object in a dark place such as at night to be recognizable, and the illumination of the second image is an illuminance or a recognizable illumination of a portion that is a saturated region in the first image. Turn off the light.

The weighting factor corresponding to the luminance value for each pixel is the first
When the luminance value of the first image is small based on the luminance value of the pixel of the image, the ratio of the luminance value of the first image is increased and the ratio of the luminance value of the second image is decreased. When the luminance value of the first image is large (saturation state), the ratio of the luminance value of the first image is reduced and the ratio of the luminance value of the second image is increased.

Here, since the illuminance of the illuminating device is increased in the first image, a region where the reflection of light emitted from the illuminating device is large tends to be saturated, and even if an object is present in the region, The first image may not be recognized. But,
The other area (unsaturated area) has recognizable brightness. On the other hand, in the second image, since the illuminance of the lighting device is weakened or turned off, the region where the reflection of the light emitted from the lighting device is large has an appropriate luminance value, and the presence of the object can be recognized. The region has low brightness and is not suitable for recognition processing.

Next, an inter-frame difference is performed using the synthesized image as an image to be processed, and a change area or an integrated area of the difference image is used as a tracking source area as a template pattern. A gray-scale pattern matching based on normalized correlation is performed using substantially the same region as a search region, a change region having a similarity higher than a predetermined value is excluded as a disturbance, and a change region having a low similarity is detected as an object to be monitored.

An image monitoring apparatus to which the method of the present invention is applied includes:
In an apparatus for monitoring an object entering a monitoring area by processing an acquired image and including an imaging unit such as an ITV camera with an illumination apparatus that captures an image of a monitoring area, an illumination state of the illumination apparatus is changed according to a field of an input image. Image input control means for controlling strong or weak;
Image input means for periodically capturing the second image and the second image due to the weak illumination state; image combining means for combining the captured first image and second image to create a processing target image; Moving object detecting means for detecting the object by performing image processing based on the processing target image. The image input control means may control the exposure state of the imaging means together with the illumination state.

According to the present invention, based on the luminance value of the normally exposed image, the weight of the luminance value of the normally exposed image is increased in the non-saturated region of the normally exposed image, and the saturated region of the normally exposed image is increased in the high speed. Since the addition is performed while increasing the weight of the brightness value of the exposure image, a composite image having a wide dynamic range can be obtained. Further, at the time of high-speed exposure, the illuminance is darkened or turned off, so that the illumination power can be reduced as compared to the case where the illuminance is always strong, and the illuminating device can have a longer life.

The present invention described above is suitable for continuous monitoring day and night. In this case, the present invention is applied to a dark scene such as a night scene where object recognition is most difficult, and a case where illumination light from an illumination device is reflected on a high-reflectance object such as a mirror or a white wall and is saturated on an image. The operation of will be described.

As shown in FIG. 2, a surveillance scene is photographed by an ITV camera with a lighting device. When the input timing is the i-field 610, the exposure state is set to the normal exposure 611,
The illuminance 612 of the lighting device is increased, and a first image 613 is input. In the next input timing i + 1 field 615, the exposure state is set to the state of the high-speed exposure 616, and the illuminance 617 of the illumination device is weakened (or turned off).
Enter Captured first image 613 and second image 6
Using 18, an image synthesis 630 is created by adding with a weighting factor corresponding to the luminance value of each pixel. The weighting coefficient according to the luminance value of each pixel is based on the luminance value of the pixel of the first image 613, and when the luminance value of the first image 613 is small, The ratio is increased to decrease the ratio of the luminance values of the second image 618. If the luminance value of the first image 613 is large (saturated), the ratio of the luminance values of the first image 613 is decreased. Thus, the ratio of the luminance value of the second image 618 is increased.

In this example, since the first image 613 is in a strong illumination state and is under normal exposure, the light receiving area 614 of the reflected light on the mirror or the like becomes saturated. It cannot be identified in the first image 613. However, in other areas (non-saturated areas), an appropriate luminance value can be obtained by normal illumination with strong illumination and a normal shutter speed. On the other hand, the second image 618 has an appropriate luminance value even in the light receiving area 614 of the reflected light by the mirror or the like due to the extinguishing or weak illumination and the high-speed exposure with a high shutter speed. Can recognize. However, other areas are dark with low brightness.

As described above, in a strong illumination light state and at the time of normal exposure, a light receiving area of light reflected by a mirror or the like is saturated. On the other hand, when the illumination is weak and the exposure is high,
Saturation hardly occurs even in a light receiving area of light reflected by a mirror or the like. By combining a first image by normal exposure and a second image by high-speed exposure at a predetermined ratio by controlling exposure in accordance with the illumination state, a dark area at night and a light receiving area of reflected light by a mirror or the like can be obtained. , A high-sensitivity image with a wide dynamic range can be obtained for the entire scene, and even in a scene where a strong noise light is partially applied in a dark monitoring area, a moving object with high accuracy can be monitored. In the case of a dark scene where there is no light at night, a saturated area rarely exists in a normally exposed image in which illumination is turned on. Therefore, the necessity of a high-speed exposure image which is not weak or turned off is small, and the addition ratio is small.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of an image monitoring apparatus according to an embodiment of the present invention. The image monitoring apparatus according to the present embodiment includes a monitoring apparatus main body 10 and a display device 7000. Or, the monitoring device body 20, the ITV camera 100, the lighting device 200,
The display device 7000 may be used. The monitoring device body 20 may be realized by mounting an image processing board on a personal computer. The monitoring device body 10 may be realized as an illuminated intelligent camera that integrates LED lighting, a camera, an image processing board, and a personal computer.

In this embodiment, first, an LED lighting device 200 is provided.
The surveillance scene is photographed with the ITV camera 100. Image input control unit
500 increases the illuminance of the illumination device 200 for the first image, and sets a normal exposure state at a normal shutter speed;
Illumination device for second image shifted by one field from first image
Set the high-speed exposure state with a low illuminance of 200 and a high shutter speed. After inputting the first image in the i-th field, the image input unit 1000 inputs the second image in the i + 1-th field. The image synthesizing unit 2000 synthesizes the images by adding the weighted coefficients corresponding to the luminance values of the pixels using the captured first image and the second image, creates an image having a wide dynamic range, and processes the image. Make an image.

The difference image creation unit 3000 performs difference processing for each pixel between frames of the processing target image using the wide dynamic range image synthesized by the synthesis image unit 2000 as a processing target image. The change area extraction unit 4000 extracts a change area by binarizing the difference image. The object detection unit 50000 includes the change area extraction unit 4
The neighborhood of the change area extracted in step 000 is combined into one integrated area, and object determination processing is performed for each integrated area by performing gray-scale pattern matching by normalized correlation processing between the immediately preceding frame and the current frame. The display control unit 6000 stores the detected image of the detected object and, in real time, displays information such as a detection position and information such as a detection date and time in response to a display request from a monitoring center or the like (not shown). The information is displayed on the display device 7000.

FIG. 3 is a block diagram showing an embodiment of the image input control section 500. Image input timing setting unit 510
Sets the image input timing to the i field, the exposure state setting unit 520 sets the exposure state to normal exposure,
The illumination intensity setting unit 530 increases the output current and
And the processing of the image input unit 1000 is performed. Next, when the image input timing setting section 510 sets the i + 1 field, the exposure state setting section 520 sets the exposure state to high-speed exposure, and the illumination intensity setting section 530 reduces the output current to reduce the illuminance of the illumination device such as an LED. The image input unit 1000
Is performed. Thus, a normal exposure image with strong LED illumination and a high-speed exposure image with weak LED illumination shifted by one field are captured.

Note that the image input control unit 500 of this embodiment controls only the illumination intensity, and exposure control is performed by a normal camera auto. In this camera auto, the normal exposure cycle is about 1/30 second, and the high-speed exposure cycle is 1/500 to 1/1000
Set to seconds.

FIG. 4 is a block diagram showing an embodiment of the inside of the image input unit 1000. When the image input control unit 500 increases the illuminance of the illuminating device 200 in the i-th field and sets it to the normal exposure state, the A / D conversion unit 1010 performs A / D conversion, and the normal exposure image input unit 1020 outputs the normal exposure image. Take in. On the other hand, when the image input control unit 500 weakens the illuminance of the illuminating device 200 in the (i + 1) th field and sets it to the high-speed exposure state, the A / D conversion unit 1010 performs A / D conversion, and the high-speed exposure image input unit 1030 Capture the exposure image.

FIG. 5 is a time chart showing image capture and composite image processing. Here, an example will be described in which the monitoring device is provided with a controllable lighting device and can be monitored continuously for 24 hours. One cycle 1510 of image processing (cycle from image capture to object detection processing) at time elapse 1500 t is k seconds. K is 100mse for online processing
c to several seconds. At this k-second interval, the illumination is turned on to capture an image, and the illumination is turned off except when capturing.

The captured image includes a normal exposure image having a period of about 1/30 second and a high-speed exposure image having a period of 1/500 to 1/1000 second. When the normal exposure image has a saturated region due to external light, the high-speed exposure image only needs to have an illuminance that optimally captures the saturated region, so that illumination is usually unnecessary. Therefore, the lighting is turned on only at the time of capturing the normal image 1540, and thereafter the lighting is turned off. The high-speed exposure image used for the composite image is captured at the timing when the afterglow due to illumination has disappeared (indicated by a circle).

FIG. 6 is a block diagram showing an embodiment of the image synthesizing unit 2000. First, the pixel luminance value calculation unit 2100
The image in the normal exposure state is captured, and the luminance value of each pixel is calculated. Normal exposure image composition ratio calculation unit 2200
Calculates the ratio of normal exposure by increasing the ratio if the luminance is low and decreasing the ratio if the luminance is high according to the calculated luminance value of the pixel (i, j). Further, assuming that the entire combination ratio is 100, the high-speed exposure image combination ratio calculation unit 2300 calculates the remainder obtained by subtracting the ratio calculated by the normal exposure image combination ratio calculation unit 2200 from 100 as the high-speed exposure ratio. That is, if the luminance value of the pixel (i, j) calculated by the pixel luminance value calculation unit 2100 is low, the ratio is decreased, and if the luminance value is high, the ratio is increased. The image addition unit 2400 calculates the ratio calculated by the normal exposure image synthesis ratio calculation unit 2200 for the pixel (i, j) to the luminance value of each pixel calculated by the pixel luminance value calculation unit 2100. , High-speed exposure image synthesis ratio calculation unit 2300
The ratio calculated in the step is calculated, and addition is performed for each pixel to create a composite image.

FIG. 7 is a flowchart showing an example of the processing procedure of the image synthesizing unit 2000. First, the ratio coefficient α of the normally exposed image is set (s101). Next, the luminance values of the normal exposure image and the high-speed exposure image are calculated for the pixel (i, j) (s102). Next, the pixels (i,
The normal image ratio luminance value (Nnode) is calculated by multiplying the luminance value of j) by the ratio coefficient α (s103). Next, the luminance value of the pixel (i, j) of the high-speed exposure image is added to the ratio coefficient (1
-Α) to calculate a high-speed image ratio luminance value (Hnode) (s104). Then, the normal image specific luminance value (Nnod
e) and the high-speed image ratio luminance value (Hnode) are added to obtain the luminance value of the pixel (i, j) of the composite image (s105), and the above processing is repeated for all the pixels (s106). In step s105, Nnode and Hnode are used to reduce the error.
Are multiplied by a coefficient raised to the power of 2n, added, and after addition, divided by a coefficient raised to the power of 2n.

FIG. 8 is an explanatory diagram showing an example of the combination ratio. In this example, for a normally exposed image brightness value 2410,
The high-speed exposure image composition ratio 2420 becomes a straight line 2430, and the gradient of this straight line becomes the ratio coefficient 1-α. The normal exposure image composition ratio is a value α obtained by subtracting the high-speed exposure image composition ratio from 100%, that is, the sum of both is 100%. here,
Normal exposure image luminance value (256 gradations) is lowest (0)
In the case of, the ratio of normal exposure is 100%, the highest (255)
In this case, the ratio is about 70%, and the remainder is the ratio of high-speed exposure. As a result, when the luminance value of the pixel of the image captured by the standard normal exposure is low, the ratio of normal exposure increases, and when the luminance value is high, the ratio of high-speed exposure increases, so that the dynamic range is wide and sharp. Images can be obtained.

FIG. 9 is an explanatory diagram showing one principle by which an image with a wide dynamic range can be obtained by the image synthesizing unit 2000. Here, a normal exposure image 2460 and a high-speed exposure image 2470 are used as images, and an example in which the output is 8 bits will be described. The normal exposure image 2460 has a luminance value of 255 in a range Isn2465 saturated by normal exposure. High-speed exposure image 2470
Now, when the high-speed exposure image range is expanded n times, the range becomes Ish 2475, which is n times the normal saturation, and the luminance value of the range 2475 is also 255. In a high-speed exposure image, the sensitivity in a darker area is lower than that in a normal exposure image, and the sensitivity in a brighter area is increased.

Here, the ratio coefficient α of normal exposure (0 ≦ 0)
α ≦ 1), the luminance value of the pixel (i, j) of the normal exposure image is Nnode, and the luminance value of the pixel (i, j) of the high-speed exposure image is Hnod.
Assuming e, the luminance value Bnode of the pixel (i, j) of the composite image can be calculated by Equation 1.

[0029]

(Equation 1)

From the equation 1, the range Isb2485 of the saturation of the synthesized image is obtained.
Is a range between the normal saturation range Isn2465 and the high-speed saturation range Ish2475, and it can be seen that a synthesized image with a wide dynamic range can be obtained.

FIG. 10 is a block diagram showing an embodiment of the difference image creation unit 3000. The noise reduction unit 3100 performs a smoothing process, a median filter process, and the like on the processing target image synthesized by the image synthesis unit 2000 to remove noise and the like. An inter-frame difference is performed for each pixel to create a difference image.

FIG. 11 is a block diagram showing an embodiment of the change area extraction unit 4000. The luminance frequency distribution calculation unit 4100 includes:
The difference image creation unit 3000 calculates the luminance frequency distribution of the difference image created by performing the difference for each pixel between frames. The binarization threshold calculation unit 4500 calculates a binarization threshold of the difference image from the luminance frequency distribution calculated by the luminance frequency distribution calculation unit 4100 by a general-purpose method such as a mode method. Binary image creation unit 46
In step 00, a binary image is created by performing a binarization process using the threshold value calculated by the binarization threshold value calculation unit 4500. Small area removal unit 48
In step 00, a binary image of a change area is created by performing filtering, expansion / contraction, or the like that is inappropriate for the detection target and excludes noise.

FIG. 12 is a block diagram showing an embodiment of the object detection unit 5000. The extraction region integration unit 5100 creates a label image of the binary image created by the change region extraction unit 4000, determines whether the distance between the labels is within a predetermined distance, and within a predetermined distance and within a predetermined distance. It is determined whether or not the circumscribed rectangle of the label group has a predetermined size, and the label groups within the predetermined size are integrated. The circumscribed rectangle calculation unit 5300 calculates a circumscribed rectangle area of the integrated label group. The circumscribed rectangular area similarity calculation unit 5500 uses the circumscribed rectangular area of one of the images as a template pattern between the images of the immediately preceding frame and the current frame, and sets the other image at substantially the same position as the circumscribed rectangular area (extended image). Normalized correlation is performed with a size of ± 1 pixel or less. The calculation of the similarity in the normalized correlation is based on the normalized correlation processing of Expression 2.

[0034]

(Equation 2)

That is, a brightness difference is obtained by normalizing the brightness of the registered template pattern and the target image (application of the light and shade pattern matching processing of the 3F-8 car number recognition system,
IPSJ 49th National Convention, late 1994), and calculates the similarity by executing the operation of Expression 1 over the entire matching region. The object determination unit 5700 determines that the object is a moving object when the similarity calculated by Expression 1 is equal to or less than a predetermined value, and otherwise excludes the object as a disturbance.

FIG. 13 is a flowchart showing an example of a processing procedure for integrating the extraction areas by the extraction area integration unit 5100. In step s210, labeling of the binary image is performed, and 1 to L
When labels up to n are given, Ln becomes the total number of labels.
In order to repeat the total label number Ln, the label number to be processed is initialized to (i) in step s220. Next, in step s230, the label number is increased by one (i is increased by +
1) Then, in step s240, it is checked whether or not all the labels have been completed. If the processing is not to be ended, the processing after step s250 is performed on the i-th label.

First, in step s250, the barycenter (center) coordinates are calculated for the i-th label. Step s2
At 60, it is checked whether the distance between the center of gravity (center) in the X direction and the Y direction between the ith label and the (i + 1) th to Lnth labels is within an allowable range. If not, step s
Return to 230. If within the allowable range, step s270
Then, it is checked whether or not the size of the circumscribed rectangle of the label within the allowable range is within a predetermined size range. If the circumscribed rectangle is not within the allowable range, a step s is performed for a new integrated region generation process.
Return to 230. If within the allowable range, step s280
Then, all the labels within the allowable range are added to the i-th label, and the added label is deleted. When the i-th to Ln-th labels are sorted in ascending order in step s290,
Returning to step s230, an integrated area generation process is performed from the new i-th. As a result, the label groups within the allowable range of the distance and the circumscribed rectangle are successively integrated.

Here, the allowable range of the distance between the X direction and the Y direction is, for example, when the object is a vertically long person, the Y range is more than the X direction.
The distance is set to be larger or longer in the direction, or if the object is long horizontally, the distance in the X direction is set to be larger than the distance in the Y direction. The same applies to the permissible range of the circumscribed rectangle. In any case, the range may be appropriately set so as to follow the detection target object depending on the range to be integrated.

FIG. 14 is an explanatory diagram showing an example of a procedure for calculating a circumscribed rectangle by the circumscribed rectangle calculation unit 5300. Since the label group integrated by the extraction area integration unit 5100 has the same label number, if only the label group is extracted, the label a5310,
Label b5320, Label c5330, Label d5340, Label e
5350, label f5360, and label g5370 are extracted. For these labels, the bounding rectangle is
Is calculated as

FIG. 15 shows a circumscribed rectangular area similarity calculation unit 5500.
FIG. 8 is an explanatory diagram showing an example of a case where a moving object is determined based on the similarity obtained by the normalized correlation processing in FIG. Bounding rectangular area 5400
Is a change area extracted by the inter-frame difference, and is a change area (moving object area) in the current frame image. The moving object does not exist inside the circumscribed rectangle 5400 in the immediately preceding frame. That is, the circumscribed rectangular area 5400 in the immediately preceding frame 5510 is a background scene where no moving object exists, and the circumscribed rectangular area 5400 in the current frame image 5530.
Is a scene including a moving object. When the correlation is performed by comparing the circumscribed rectangular area 5400 between the immediately preceding frame image and the current frame image, the background is mostly hidden, and the similarity is considerably reduced. Thereby, the presence or absence of the moving object can be accurately determined.

FIG. 16 shows a circumscribed rectangular area similarity calculation unit 5500.
It is a flowchart which shows an example of the processing procedure of. Step s3
Reference numeral 10 registers, as a template pattern, a grayscale image of a circumscribed rectangular area calculated by the integrated area circumscribed rectangle calculation unit with respect to the current input image. In step s320, a circumscribed rectangular area at almost the same position as in step s310 is set as a pattern matching area for the immediately preceding frame image. A step s330 performs a light and shade pattern matching with the pattern registered in the step s310 to calculate a similarity. Calculation of the similarity is based on Equation 2. Step s340 is:
It is determined whether the calculated similarity is equal to or greater than a predetermined value. If the value is equal to or larger than the predetermined value, step s350 is determined to be a moving object because it is not similar to the background. If it is less than the predetermined value, step s360 is determined to be a disturbance because it is similar to the background. A step s370 decides whether or not all of the generated circumscribed rectangular areas have been completed.
Return to 310.

FIG. 17 is an explanatory diagram showing an example in which the detection result is displayed on the display device 7000. When a person 6200 is detected,
The display control unit 6000 uses the stored circumscribed rectangular position of the detected object, and superimposes the current input image on the display device 7000 to form a circumscribed rectangular frame.
Display control of 6300. In FIG. 17, the display control unit 6000
However, when display control of the detected object is performed on the display device 7000, any display method can be used as long as the detected object can be clearly indicated. By displaying the information on the display device 7000 in this manner, for example, in the case of monitoring a person, the observer can display the person and the detection information on the display device 70.
You can grasp online on 00. If the display device 7000 is located at a remote place, the fact that a person has been detected by a standard communication means such as RS-232C may be notified to a videophone or the like, and displayed on a remote display device.

[0043]

According to the present invention, first, an IT with a lighting device
When shooting a surveillance scene with a V camera, a normal exposure state in which the illumination is increased is set at the input timing of the i-field, and the first image is input. At the input timing of the I + 1 field, the illumination is not weak and is turned off in the high-speed exposure state. The second image is input after setting, and the two images are added with a weighting factor corresponding to the luminance value of each pixel to create an image composite, so that a monitoring image with a wide dynamic range can be created, and a dark scene such as at night can be created. Even when strong noise light such as a headlight hits the monitoring, high-sensitivity monitoring becomes possible. Also,
Since illumination is performed at the time of capturing a normally exposed image, current consumption can be reduced, and the life of the illumination device can be extended.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of an image monitoring apparatus according to the present invention.

FIG. 2 is an explanatory view showing the operation of the present invention.

FIG. 3 is a block diagram showing an embodiment of an image input control unit.

FIG. 4 is a block diagram showing an embodiment of an image input unit.

FIG. 5 is a time chart showing timings of capturing an image and lighting.

FIG. 6 is a block diagram showing an embodiment of an image synthesizing unit.

FIG. 7 is a flowchart illustrating an example of a processing procedure of an image combining unit.

FIG. 8 is an explanatory diagram illustrating an example of a composition ratio of an image composition unit.

FIG. 9 is an explanatory diagram showing the principle of expanding a dynamic range by image synthesis.

FIG. 10 is a block diagram showing an embodiment of a difference image creation unit.

FIG. 11 is a block diagram showing an embodiment of a change area extraction unit.

FIG. 12 is a block diagram illustrating an embodiment of an object detection unit.

FIG. 13 is a flowchart illustrating an example of a procedure for integrating the extraction regions of the object detection unit.

FIG. 14 is an explanatory diagram showing one procedure of calculating a circumscribed rectangle by a circumscribed rectangle calculation unit.

FIG. 15 is an explanatory diagram showing an example of moving object determination by normalized correlation processing.

FIG. 16 is a flowchart illustrating an example of a processing procedure of a circumscribed rectangular area similarity calculation unit.

FIG. 17 is an explanatory diagram illustrating an example of a detection result displayed on a display device.

[Explanation of symbols]

100 ITV camera, 200 lighting device, 500 image input control unit, 510 image input timing setting unit, 520 exposure state setting unit, 530 illumination intensity setting unit, 1000 image input unit, 1010
... A / D converter, 1020 ... Normal exposure image input unit, 1030
... High-speed exposure image input unit, 2000 ... Image synthesis unit, 3000 ... Difference image creation unit, 4000 ... Change area extraction unit, 5000 ... Object detection unit, 6000 ... Display control unit, 7000 ... Display device.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G03B 15/05 G03B 15/05 5C076 G06T 1/00 430 G06T 1/00 430G 5C083 3/00 300 3/00 300 5L096 7/20 7/20 A G08B 13/196 G08B 13/196 H04N 1/387 H04N 1/387 5/225 5/225 C (72) Inventor Yoshiki Kobayashi 7-1-1, Omikamachi, Hitachi City, Ibaraki Prefecture No. within Hitachi, Ltd.Hitachi Research Laboratories (72) Inventor, Hisao Ota 1410 Inada, Hitachi, Ibaraki, Japan Digital Media Products Division (72) Inventor, Shigehisa Sakimura Omikamachi, Hitachi, Ibaraki, Japan No. 1-1, Hitachi Research Laboratory, Hitachi Ltd. (72) Inventor Yoshiro Sakuma Shimo, Tokyo F-term (reference) 2H053 AA01 AB03 AD00 AD21 BA71 BA91 5B047 AA07 AB02 BB06 CA19 5B057 AA19 BA02 CA08 CA13 CB08 CB12 CC01 CE08 CH08 DA02 DA16 DB03 DB09 DC32 5C022 AB15 AC69 A05 CB03 CC03 CH02 EA01 FE12 HA19 5C076 AA12 AA40 BA06 5C084 AA02 AA07 AA08 AA13 BB27 BB32 CC16 CC19 DD12 DD66 GG20 GG39 GG42 GG43 GG44 GG52 GG56 GG57 GG61 GG68 GG73 CA17 AGG02 CA17A17A

Claims (5)

[Claims] 1. A method of monitoring an object that enters a monitoring area by performing image processing on a difference image between input images obtained by capturing the monitoring area, comprising: Capture, and in the next field, turn on or off the illumination and capture a second image, capturing the first image and the second image.
An image monitoring method, comprising: creating a composite image based on the image of (i), and generating the difference image by an inter-frame difference using the composite image as a processing target image. 2. The illumination of the first image according to claim 1, wherein the illumination of the first image is an illuminance capable of recognizing an object in a dark place such as at night, and the illumination of the second image is a saturated area in the first image. An image monitoring method, characterized in that a part is set to an illuminance or a lighting-off state that can be recognized. 3. The image according to claim 1, wherein the area where the luminance value of the first image is unsaturated is based on the luminance value of the pixel of the first image. The weight of the luminance value is increased, while the area where the luminance value of the first image is saturated is created by adding the luminance value of each pixel with a weighting factor obtained by increasing the weight of the luminance value of the second image. Characteristic image monitoring method. 4. An apparatus for processing a captured image to monitor an object entering a monitoring area, comprising: an imaging device with an illumination device for capturing an image of a monitoring area, wherein an illumination state of the illumination apparatus is determined according to a field of an input image. Input control means for controlling strongly or weakly, image input means for periodically capturing a first image in a strong illumination state and a second image in a weak illumination state, and a first and a second captured image. An image monitoring apparatus comprising: an image synthesizing unit that synthesizes an image to generate a processing target image; and a moving object detection unit that detects an object by performing image processing based on a processing target image between frames. 5. The image monitoring apparatus according to claim 4, wherein the image input control means controls an exposure state of the imaging means together with the illumination state.