JPH11112966A - Moving object detecting device, moving object detecting method, and computer-readable storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a moving object of a desired size within a wide range with respect to a distance of the object, independently of the magnification of photographing in a system for monitoring the moving object through the use of a video camera. SOLUTION: After entering the magnification of a camera and the size of a moving object to be detected (steps S1, S2), the moving object is detected from a image picked up video image, and a distance up to the moving object is measured (steps S3, S4). Then the size of the detected moving object is detected, the detected size is compared with the size of the moving body based on a reference distance and a reference magnification obtained in advance. Then whether or not the detected mobile body is a moving body of a desired size is decided based on the comparison result, and when the mobile object is a mobile object of the desired size, an alarm is sounded (steps S6, S7).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and a method for detecting a moving object having a desired size from an image picked up by an image pickup apparatus such as a video camera, and a computer-readable storage medium used for the method.

[0002]

2. Description of the Related Art Conventionally, there has been known a moving object detecting apparatus which detects an intruding object by detecting a moving object in an image being picked up by a video camera for monitoring purposes or the like. In such an apparatus, the size of a moving object to be detected is often known in advance, and detection of only a moving object having a specific size has already been attempted.

[0003]

However, in an image captured by an image capturing apparatus such as a video camera, even if the same object is captured, different sizes are captured depending on the distance to the subject and the photographing magnification.

This situation will be described with reference to FIG. FIG.
In 6, reference numeral 120 denotes an imaging center, and 127 denotes an optical axis of the imaging system. Reference numerals 121, 122, and 123 represent planes orthogonal to the optical axis 127, and are located at distances of 1 m, 2 m, and 3 m from the imaging center 120, respectively. 124, 125,
Reference numerals 126 denote spheres each having a radius of 1/3 m, and their centers are on the surfaces 121, 122, and 123, respectively. The angle of view of this imaging system is 36.0 degrees in the horizontal direction and 27.0 degrees in the vertical direction.
In FIG. 16, the angle of view in the horizontal direction is expressed as viewed from directly above. Reference numerals 129 and 130 denote ranges of the angle of view as viewed from the imaging center 120, respectively.
The angle between 130-120-129 is 36.0 °,
127-120-129 and 127-120-130 together make an angle of 18 °. Also, the image captured at this time is composed of 640 pixels in the horizontal direction × 480 lines in the vertical direction.

FIG. 17 shows an image captured at the size of 640.times.480 pixels.
124, 125, and 126 show the sizes of the images 125 and 126, respectively, in the frame.
As can be seen from FIGS. 16 and 17, even objects of exactly the same size are displayed in the frame as different sizes depending on how far away from the imaging center. FIG. 17 shows that the sphere 124 in FIG. 16 occupies approximately 19 ° in diameter (angle formed by A-120-A ′) in the horizontal direction, and is imaged as a size of about 327 pixels on the frame. In the case of the sphere 125, about 9.5 ° (B-120
−B ′), imaged as approximately 163 pixels,
In the case of the sphere 126, the image is also taken as about 6.4 ° and about 109 pixels.

Next, FIG. 18 shows a case where the magnification is changed optically in the image pickup system. D ₁ -120-D ₁ ′ is
The reference magnification is 1.0 times at an angle of view of about 36.0 °.
D ₂ -120-D ₂ ′ is an angle of view of about 18.5 °, and indicates an angle of view when the magnification is 2.0 times. D ₃ -120
−D ₃ ′ is an angle of view of about 12.4 °, and indicates an angle of view when the magnification is 3.0 times. When the magnification is changed in this manner, the degree of the angle of view that corresponds to the frame size of 640 pixels of the captured video changes. Here, when the magnification is increased, the size of the subject is also increased in proportion to the magnification. That is, the size of the subject relative to the frame size of the video becomes large.

For this reason, conventionally, in order to detect an object of a specific size, the magnification is fixed and the distance to the subject can be effectively used only for detecting an object within a very limited range. was there.

Accordingly, an object of the present invention is to provide a device for detecting a moving object of a set size, which can deal with a wider range of a distance from a camera to a subject than a conventional device.
In addition, it is to enable detection even when the imaging magnification is variable.

[0009]

In a moving object detecting apparatus according to the present invention, moving object detecting means for detecting a moving object from image data input by an imaging means, and information on the size of the moving object detected by the moving object detecting means. Moving body size detecting means, detecting means for measuring the distance from the moving body detected by the moving body detecting means to the imaging means, distance information measured by the measuring means and detected by the moving body size detecting means There is provided a predetermined moving object detecting means for detecting a moving object having a predetermined size from information on the size of the moving object.

In the moving object detection method according to the present invention, a step of detecting a moving object from image data input by the imaging means, a step of detecting information on the size of the detected moving object, A step of measuring a distance to the imaging means; and a step of detecting a moving object having a predetermined size from the measured distance information and the information on the detected size of the moving object.

A computer-readable storage medium according to the present invention stores a program for executing the steps in the moving object detection method.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of a main part according to the first embodiment of the present invention. FIG.
, A photographing lens 12 having a zoom function, symbolically comprising a zooming lens 1 for zooming and a focusing lens 2 for focusing. The photographing lens 12 is used to convert an optical image of a subject into an image sensor 3 such as a CCD.
The imaging device 3 outputs an electric signal indicating the optical image to the camera processing unit 4. The camera process unit 4 performs well-known processing (gain correction, γ correction, color balance adjustment, and the like) on the output of the image sensor 3, and outputs a video signal of a predetermined format. Focus detection target area setting unit 7
Designates an area in the video to be automatically focused by the focusing control unit 8.

FIG. 2 shows an example of a focus detection target area on a captured image. In FIG. 2, the image area 31
Is assumed as a digital image composed of 640 pixels × 480 pixels. A rectangular area composed of 140 × 140 pixels in the video area 31 is shown as the focus detection target area 32. Here, a case where a circular object 33 is imaged is shown as an example of a main subject existing in the focus detection target area 32.

The focus detection target area 32 is determined under the control of the overall control unit 20 as relative position information on the image pickup plane as to which area on the image pickup plane is to be focused on the subject. It is set in the focus control unit 8 via the focus detection target area setting unit 7. The focus control unit 8 controls the high frequency component included in the video signal of the portion corresponding to the region set by the focus detection target region setting unit 7 in the video signal output from the camera process unit 4 to be maximized. By controlling a focusing lens motor (stepping motor) (not shown) to move and adjust the position of the focusing lens 2 in the optical axis direction, the subject is automatically focused. Here, the position of the focusing lens 2 (the position of the lens at each point in the position range in which the focusing lens can be driven by the focusing lens motor) is a pulse number indicating how many pulses the focusing lens motor has been driven from the reference position. And so on and output to the outside of the focusing control unit.

The variable power setting section 9 includes a zooming lens motor (stepping motor) (not shown) for the zoom control section 10.
Is driven to move and adjust the position of the zooming lens 1 in the optical axis direction to set a target magnification when performing zoom control. The variable power setting unit 9 sets the set magnification received from the overall control unit 20 to the zoom control unit 10 as a zooming motor drive pulse value corresponding to the magnification value as a set value for zooming lens motor control. In accordance with the set value, the zoom control unit 10 controls the zooming lens motor to move and adjust the position of the zooming lens 1 in the direction of the optical axis, thereby enabling imaging with a desired zoom. Here, the position of the zooming lens 1 is also determined by the focusing lens 2
Similarly to the above, the position in the movable range of the zooming lens 1 is output to the outside of the zoom control unit 10 in the form of the number of pulses indicating how many pulses the zooming lens motor has been driven from the reference position. It is configured so that: still,
The components described above are well known in video cameras and the like.

Next, the moving object detecting section 5 detects a moving object in an image from a video signal output from the camera processing section 4. As this kind of moving object detection method, for example, a method using a background difference is known. That is, as shown in FIG. 3, an image 41 that does not include a moving object in the observation region is previously captured and stored. Next, the monitoring image 42 (image being captured) at the time of observation is compared with the image 41, and a difference value image 43 obtained by calculating a difference between pixel values for each corresponding pixel is generated. The difference value image 43 is obtained as an image having a significant pixel value only in a portion different from the previously stored image 41 that does not include motion. An area 44 composed of pixels having a significant pixel value (a value sufficiently larger than zero) included in the obtained difference value image 43 is detected as a moving object.

The size of the detected moving object (for example,
It is determined whether or not the number of pixels included in the area 44) corresponds to a size assumed in advance, and thereby it is determined whether or not the moving object has a desired size to be detected. At this time, the moving body size correction unit 21 corrects the detected size of the moving body to a reference distance and a size detected at the reference magnification according to the subject distance from the camera to the moving body. As a result, a moving object having a specific size can be detected in a wider range than in the related art.

Next, the moving body detecting section 5 will be described in more detail with reference to FIG. In FIG. 4, a video capture unit 51 captures a video signal output from the camera process unit 4 in FIG. 1 and writes it as a digital image in a frame memory 52 for each frame. On the other hand, the background memory 53 previously captures and stores an image such as the background image 41 in FIG. 3 captured in a state where no moving object is present by the initialization circuit (not shown) before the start of monitoring.

The difference calculation unit 54 inputs a pixel value obtained by sequentially reading out the corresponding pixels from the two images held in the frame memory 52 and the background memory 53 in the sequential scanning order. Then, a value obtained by subtracting the output pixel value of the background memory 53 from the output pixel value of the output is output (however, an absolute value is output). When the difference results (absolute values) output from the difference calculation unit 54 are arranged in the scanning order for one frame, the difference value image 43 shown in FIG.
Can be obtained.

Next, the binarizing section 55 binarizes the output of the difference calculating section 54 using a predetermined threshold value which is considered to be a significant value, so that there is no significant difference between the two images. 1 (ON: black) only for the pixels in the area where
(OFF: white) binarized pixel values are sequentially output in the scanning order. Next, the noise removing unit 56 removes an isolated pixel or a minute black pixel region in the binary image caused by noise mixed in due to various factors in the processing up to this point, and removes a minute hole (a minute pixel in the black pixel connection region). A small white pixel area is removed.

FIG. 5 shows a configuration example of the noise removing unit 56.
In FIG. 5, reference numerals 601 to 609 denote latches, which hold bit data of 9 pixels (9 bits of 1 bit for each pixel) in a 3 × 3 pixel area 600 as shown in FIG. Reference numerals 61 and 62 denote FIFO memories each having 1
Data for the number of pixels on the scanning line is held. That is, FIF
O61 holds data input before the current scanning line by a scanning line, and FIFO 62 holds data input two scanning lines before the currently input scanning line.

Of the nine latches 601 to 609,
Reference numerals 601 to 603 hold bit data corresponding to three pixels arranged on the currently input scanning line, 604 to 606 are adjacent to one scanning line (adjacent in the sub-scanning direction), and 607 to 609 are further one scanning line. It holds bit data corresponding to three pixels arranged on an adjacent (adjacent in the sub-scanning direction) scanning line. This means that the data is sequentially shifted one pixel at a time in synchronism with the sequential input of the binary image data output from the binarization unit 55 to the raster scanning line.
It is possible to sequentially scan an image in a nine-pixel area of three pixels.

Reference numeral 63 denotes a ROM, and latches 601 to 6
The output 9 bits of 09 are used as an address input, and 1-bit data is output according to the state of the output 9 bits of these latches 601 to 609. The ROM 63 stores a 1-bit signal when 5 or more bits of the 9-bit address input are 1, for example.
Is output in advance, and data that outputs 0 when 5 or more bits are 0 is stored in advance. That is, 9 of 3 × 3 pixels
When five or more pixels in the pixel area are black pixels, black pixels are output, and when four or less pixels are white pixels, white pixels are output. By using the ROM 63 as a look-up table in this manner, removal of isolated pixels and the like can be realized. The noise removing unit 56 has a configuration of a pipeline processing circuit.
Although a delay corresponding to a pixel occurs, binary pixels from which noise has been removed are sequentially output in the raster scanning order.

In FIG. 4, bit map memory 57
Accumulates the binary pixel data output from the noise removing unit 56 for one frame. The moving object position detection unit 58 sequentially inputs the noise-removed binary image data in the raster scanning order as shown in FIG. 7 and obtains coordinate values indicating a rectangular area 81 surrounding the black pixel area as shown in FIG. (X _min ,
Y _min ) and (X _max , Y _max ) are detected. The detection of these coordinate values can be easily realized by a known circuit basically configured using a counter and a comparator.

That is, four counters and four buffers for detecting and holding X _min , X _max , Y _min , and Y _max respectively are prepared. First, the counter that detects X _min counts the number of pixels in the main scanning direction until the first black pixel appears in the data on each scanning line (counts a synchronization pulse in the main scanning direction, not shown). The comparator compares the count value with the value counted in the previous scan line held in the buffer for holding X _min ,
If the value of the counter is smaller than the value of the buffer, the value held in the buffer holding X _min is changed to the current count value. Otherwise, the value of the buffer holding X _min is not changed. Here, the value holding X _min is initialized for each scanning line to a value larger than the number of pixels included on the main scanning scanning line.

With respect to X _max , the pixel position on the main scanning line at the point where the pixel returns to the white pixel after detecting the black pixel on the scanning line may be detected (until the change from the black pixel to the white pixel is detected). Scan sync pulses are counted). The larger than the value of X _max so far, it updates the value of X _max, otherwise not updated. For Y _min , the number of scanning lines scanned so far (counting the sub-scanning synchronization pulse) when a scanning line including a black pixel is detected for the first time is counted, and for Y _max , the scanning pixel including a black pixel is included. After the detection of the scanning line, the number of scanning lines at the time when the scanning line not including the black pixel is detected again may be counted.

When the scanning of one frame of the binary image is completed, the coordinates (X _min , Y _min ) and (X _max , Y _max ) of the diagonal vertexes of the rectangular area surrounding the moving object can be detected. .

Next, the moving body size detecting section 6 in FIG. 1 outputs (X _min , Y _min ) from the moving body detecting section 5,
From the value of (X _max , Y _max ) and the noise-removed binary image data held in the bitmap memory 57 of FIG. 4, the size of the moving object is coefficientd.

FIG. 9 shows an example of the configuration of the moving object size detecting section 6. In FIG. 9, the scan clock generator 91 receives the coordinates (X _min , Y _min ) and (X _max , Y _max ) of the diagonal vertices of the rectangular area surrounding the moving object output from the moving object detector 5. Then, addresses for accessing only the rectangular area in the bitmap memory 92 are sequentially generated (in a raster scanning format).

That is, in the main scanning direction, X _min to X _max
Up to (X _max −X _min +1) pixels in the sub-scanning direction,
A scan clock for (Y _max -Y _min +1) scan lines from Y _min to Y _max is generated, and the area indicated by 81 in FIG. 8A is indicated by 82 in FIG. 8B. is (X _max -X _min +1), is output by the (Y _max -Y _min +1) bit map memory 92 as binary image of the image _{(X max -X min +1) ×} (Y max -Y min +1)
Only the pixels in the black pixel portion of the binary image of the pixels (that is, the black pixel is output as the pixel value 1 and the white pixel is output as the pixel value 0), and only the pixel value 1 output in this raster scanning format is output. ) By counting, the area of the extracted moving object is coefficiented as the number of black pixels. The number of black pixels (area) is used as the moving object size.

The initialization / read-out section 93 initializes the scan clock generation section 91 and the counter 94 under the control of the overall control section 20 in FIG. Output to the unit 20.

Next, the distance measuring section 11 of FIG. 1 will be described. The distance measuring unit 11 outputs the number of driving pulses of the focusing lens motor that indicates the position of the focusing lens 2 output from the focusing control unit 8 (indicates how many pulses the focusing lens motor has been driven from the reference position to the current position). The number of pulses of the zooming lens motor output from the zoom control unit 10 indicating the position of the zooming lens 1 (the number of pulses indicating how many pulses of the focusing lens motor have been driven from the reference position to the current position) ) And outputs the distance from the camera to the subject on which the camera is in focus.

The imaging lens 12 having the zoom function shown in FIG.
The lens having the focusing lens 2 on the imaging surface side of the imaging element 3 and the zooming lens 1 on the object side is called a rear focus lens. In the rear focus lens, when the position of the zooming lens 1 is changed, the focal position also moves. Therefore, if the focusing lens 2 is not moved, a focused image cannot be obtained.

Therefore, in the rear focus lens 1, the position of the focusing lens 2 (that is, the number of pulses of the zooming lens) (that is, the number of pulses of the zooming lens) is changed.
When the number of pulses of the focusing lens motor is changed to various values, the distance from the camera to the in-focus subject at each position is measured and obtained in advance. Then, the position of the zooming lens 1 (the number of zooming motor drive pulses required for movement from the reference position) and the position of the focusing lens 2 (the number of focusing lens motor drive pulses required for movement from the reference position) are input as addresses. A look-up table for outputting the distance from the camera to the object to be focused at that time as data is constituted by a ROM.

Here, for example, assuming that the range of possible values of the number of driving pulses of the zooming motor and the number of driving pulses of the focusing motor is 0 to 2047, the memory space is 2K × 2K = 4M (2 ¹¹ × 2 ¹¹ = 22
²² ), the data dynamic range is set to 8 bits. In other words, if the roughness of the distance measurement is 256 and the range of 0 mm to ∞ is represented by 256 different distances, it can be constituted by a ROM having a capacity of 4 MByte. If necessary, the dynamic range of the data may be 16 bits or the like. In this case, the focusing distance is expressed as any of 65536 different distances in the range of 0 mm to ∞.

FIG. 10 shows an example of the configuration of the overall control unit 20. In FIG. 10, reference numeral 22 denotes a CPU; 23, a ROM storing a program as a storage medium according to the present invention;
Is a RAM, 25 to 29 are I / O ports, 39 is a communication interface, and 30 is a bus. The CPU 22 uses the RO
The program stored in M23 is read out and operates according to the procedure. In the course of the operation, information that needs to be temporarily stored and information that changes depending on the situation are stored in the RAM 24.
Hold on. Note that a semiconductor memory, an optical disk, a magneto-optical disk, a magnetic medium, or the like may be used as the storage medium.

The I / O port 25 serves as an interface between the CPU 22 and the moving object detecting section 5. The I / O port 26 serves as an interface between the CPU 22 and the moving body size detection unit 6. The I / O port 27 is a CPU
The I / O port 28 is an interface between the CPU 22 and the focus detection target area setting unit 7.
Interface with the I / O port 29
2 and an interface with the magnification setting unit 9.
The communication interface 39 communicates with an external device. For example, when the size of a moving object to be detected is input from an external device, or when a moving object having a desired size is detected, the fact is transmitted to the outside. Used to notify.

Next, using the flowchart shown in FIG. 11, the program procedure stored in the ROM 23 is read, and the program is executed by the CPU 22 to detect a moving object whose size (size) is known in advance. A series of operations in such a case will be described. In FIG. 11, when the process is started, in step S1, the communication interface 39 is started.
And input a desired scaling factor from an external host computer through the I / O-5 (29) of FIG. 10, and set the scaling factor D input to the scaling setting unit 9 via the I / O-5 (29) in FIG. As a result, the magnification setting unit 9 controls the zoom control unit 10 to control the zooming lens motor in accordance with the magnification value D, and sets the apparatus to a desired magnification D as described above.

After step S1, the process proceeds to step S2, where the size S of the moving object to be detected is input from an external host computer via the communication interface 39. The input size information S is held in a predetermined area on the RAM 24. Here, the size S of the moving object is the reference distance (1 in the present embodiment) of the camera used in the present system.
m) is set to the reference magnification (in the present embodiment, the reference magnification is set to 1.0 × when the horizontal angle of view is 36 ° and the vertical angle of view is 27 °). The number of pixels at the corner (124 in FIG.
Are input (approximately 84,000 pixels).

Next, the process proceeds to step S3, and enters a moving object detection loop of a desired size (steps S3 to S6). FIG. 12 shows the details of step S3. In step S3, first, the moving body detection unit 5 is accessed via the I / O-1 (25) in step S30, and the diagonal vertex coordinates (X) of the rectangular area 81 surrounding the moving body output from the moving body position detection unit 58 are output. _min ,
Y _min ) and (X _max , Y _max ). Then, the process proceeds to a step S31, at which (X _min , Y _min ), (X
_max , Y _max ), the rectangular area 81 surrounding the moving object
Coordinate values of the center point, (X _c, Y _c) and obtained by calculation of _{X c = (X max -X min} ) / 2 Y c = (Y max -Y min) / 2.

Then, the process proceeds to a step S32, and a step S3
The coordinate values (X _c , Y _c ) of the center point of the rectangular area surrounding the moving body obtained in 1 are passed through the I / O-3 (27) in FIG. X _c and Y _c are set in the focus control unit 8. Thereby, the distance measuring unit 1
The target for measuring the distance to the focused object in step 1 is set. After a series of processes in step S3 is completed, FIG.
The process returns to the routine of Step 1 and proceeds to Step S4. In step S4, the distance to the moving object is detected.

FIG. 13 shows the details of step S4. In step S4, first, in step S40, the focus control unit 8 fetches a signal indicating whether or not the camera is in focus via the focus detection target area setting unit 7, and determines whether or not the camera is in focus. Judge. If it is determined that the camera is not in focus, the procedure of step S40 is repeated again. If it is determined that the camera is in focus, the process proceeds to step S41. In step S41, the distance information Lo to the in-focus subject, which is the output of the distance measuring unit 11, is read via the I / O-4 (28) in FIG. 10, and the process proceeds to step S42. As described above, Lo is a code representing the distance from 0 mm to ∞ as an 8-bit or 16-bit code. Therefore, in step S42, focusing is performed by the camera based on a correspondence table (not shown) registered in the program in advance. Decoding is performed to the distance L (unit: meter) to the middle object.

After a series of processes in step S4 is completed, the process returns to the routine in FIG. 11 and proceeds to step S5.
FIG. 14 shows the details of step S5. In step S5, first, in step S50, I / O-2 (2
6), the moving body detection unit 6 is accessed, and the moving body size (number of pixels) So is input. Then, the process proceeds to a step S51, wherein the scaling factor D input in the step S1 and the step S4
The distance L from the camera obtained in step 2 to the focused object;
A reference distance of 1.0 m from So input in step S50
And the size S ′ that would be obtained when the image is taken at the reference magnification is obtained by the calculation of S ′ = So × (L / D) ² .

Then, the series of processes in step S5 is completed, the process returns to the routine in FIG. 11, and proceeds to step S6. In step S6, it is determined whether or not the ratio between the size S 'calculated in step S51 and the size S of the moving object to be detected input in step S2 and stored in the RAM 24 is within a predetermined value. That is, it is determined whether or not 0.8 <S '/ S <1.2 to determine whether or not the moving object has a desired size.

If the size is not the desired size, that is, if 0.8.gtoreq.S '/ S or 1.2.ltoreq.S' / S, the process returns to step S3, and the procedure after detecting the moving object is repeated again.

If the size is the desired size, that is, if 0.8 <S '/ S <1.2, the process proceeds to step S7. In step S7, the fact that a moving object having a desired size has been detected is notified to an external device via the communication interface 39. A series of procedures for detecting the moving object having the determined size is ended.

The above-mentioned constants 0.8 and 1.2 are not necessarily limited to these, but may be adjusted according to the equipment used, the use environment, etc., for example, 0.75, 0.25 or 0.8.
5, 1.15 or the like.

Next, a second embodiment will be described.
In the above-described first embodiment, the moving object size input from step S2 in the flowchart of FIG. 11 is not necessarily a position detected by a reference distance from the camera and the angle of view of the camera is also a reference magnification. The present invention is not limited to the number of pixels of a subject to be a moving object. In other words, during shooting when capturing a moving subject as a moving body size,
The horizontal size X of the rectangular area surrounding the subject
A desired size consisting of w (pixel) and vertical size Yw (pixel) may be input.

In this case, what is corrected in step S5 is not the moving object size So obtained by the moving object size detection unit 6, but information relating to a rectangular area surrounding the moving object obtained by the moving object detection unit 5. That is, the details of step S5 are changed to the flowchart shown in FIG.

In FIG. 15, in step S50a,
The moving object input from the moving object detection unit 5 in step S30 is
Diagonal vertex coordinates (X_min, Y_min),
(X _max, Y_max), Xo = X_max-X_min Yo = Y_max-Y_min And calculate the horizontal width Xo and the vertical direction of the rectangular area.
Direction height Yo.

Then, the process proceeds to a step S51a, wherein the scaling factor D input in the step S1, the distance L from the camera to the focused object obtained in the step S42, and the step S5
From Xo and Yo obtained in 0a, the horizontal size Xo ′ of the rectangle surrounding the subject and the vertical size Yo ′ that would be obtained when photographing at the reference distance of 1.0 m and at the reference magnification are Xo. '= Xo × (L / D) Yo ′ = Yo × (L / D) Then, the series of processes in step S5 is completed, and the process returns to the routine in FIG.
Proceed to.

In the present embodiment, step S6 also corresponds to Xw and Yw input in step S2, and to Xw and Yw calculated in step S51a, respectively.
Xo ', Yo' calculated in step S51a, respectively.
And it is determined whether the ratio is within a predetermined value. That is, by determining whether 0.8 <Xo '/ Xw <1.2 and 0.8 <Yo / Yw <1.2, it is determined whether the moving object has a desired size. . Xo ', Yo'
Are satisfied, it is determined that a moving object of a desired size has been detected, and otherwise, a change is made so as not to be detected.

The constants such as 0.8 and 1.2 are not limited to the above, and are, for example, 0.85 and 1.15 and 0.
75, 1.25, etc.

According to the present embodiment, the moving object size detector 6 does not necessarily need to count the number of pixels occupied in the moving object image, so that the circuit configuration can be simplified.

Next, a third embodiment will be described.
The moving object size input in step S2 of the flowchart in FIG. 11 is not necessarily limited to the formats disclosed in the first and second embodiments. That is, as the moving body size,
The actual size of the subject to be detected may be input.

In this case, the actual size is, for example, a vertical length (height) H (m) and a horizontal length (width) W (m) when viewed from the front.
Should be input as At this time, if the angle of view is 36.0 ° and the distance from the camera to the subject is 1 m, the horizontal width in the captured image is about 0.65 m, and this is input as an image with 640 pixels. Will be. Therefore, the horizontal length (width) W (m) and the horizontal size Xw of the rectangle surrounding the subject under the conditions of the reference distance and the reference angle of view (magnification) described in the second embodiment are used. The relationship is obtained as Xw = (W / 0.65) × 640.

The vertical angle of view of the camera is 27.0 °.
Is a width of about 0.48 in the image captured under the reference distance.
m, and the image is input as 480 pixels, so that the vertical size Yw of the rectangle surrounding the subject under the conditions of the reference distance and the reference angle of view (magnification) described in the second embodiment is Yw = (H / 0.48) × 480.

As described above, if the moving body size is input as the actual size in step S2 and converted into Xw and Yw based on the above equations, the operation can be performed in exactly the same manner as in the second embodiment.

According to the present embodiment, the size of the moving object can be input without being influenced by the specifications of the camera system, and the usability can be improved.

Next, a fourth embodiment will be described.
A range of a certain value may be input as the desired moving object size in the above-described second embodiment. That is, as Xw _min ≦ Xw ≦ Xw _max, if the value of Xw from X _min to the value of Xw _max,
It is assumed that it is in a desired size. The same applies to Yw.

At this time, in step S6, 0.8 <Y '/ Yw _max and Yo' / Yw _min <
1.2 and 0.8 <Xo '/ Xw _max and Xo' / Xw _min <
1.2, it is assumed that the moving object has a desired size.

That is, in the first embodiment, S _min <S <S _max is satisfied, and it is determined whether 0.8 <S ′ / S _max and S ′ / S _min <1.2 are satisfied. Become.

[0063] In the third _{embodiment, H min ≦ H ≦ H max} W min ≦ W ≦ W max of _{H min, H max, W min} , enter a _{W max, Yw max = (H} max / 0.48) × 480 Yw _min = (H _min /0.48)×480 Xw _max = (W _max /0.65)×640 Xw _min = (W _min /0.65)×640 Modifications similar to those of the embodiment may be made.

According to the present embodiment, it is possible to cope with a case in which the size of the moving object changes depending on the elastic and easily deformable moving object or the direction in which the image is taken.

In each of the above embodiments, an example in which the correction based on the magnification and the distance to the moving object is the size of the detected moving object has been described. However, the present invention is not limited to this. Of course, the present invention may be implemented in a form in which information relating to the size of a moving object to be detected as information is corrected.

In each embodiment, as shown in the flow chart of FIG. 11, when a moving object of a desired size is detected and an alarm is issued, a series of processing is terminated. However, the moving object detection may be repeated again. That is, after step S7 of FIG. 11 is completed, the process may return to step S3 again.

[0067]

As described above, according to the present invention,
By using the information on the distance to the moving object detected from the video, the information on the size of the moving object to be detected, and the information on the size of the moving object detected from the image, a specific monitoring area can be specified in a wider monitoring area than before. A moving object of a size can be detected. Further, a moving object of a specific size can be detected even when the imaging magnification is variable. Further, by utilizing the information of the focus control for the distance measurement, the above effect can be obtained with a simpler configuration.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a configuration diagram illustrating an example of a focus detection target area.

FIG. 3 is a configuration diagram illustrating a moving object detection direction based on a background difference.

FIG. 4 is a block diagram illustrating a configuration of a moving object detection unit.

FIG. 5 is a block diagram illustrating a configuration example of a noise removing unit.

FIG. 6 is a configuration diagram showing a 3 × 3 9-pixel area.

FIG. 7 shows a noise-removed 2 input in a tester scanning order.
FIG. 3 is a configuration diagram illustrating value image data.

FIG. 8 is a configuration diagram showing a rectangular area surrounding a pixel area.

FIG. 9 is a block diagram of a moving object size detection unit.

FIG. 10 is a block diagram illustrating a configuration example of an overall control unit.

FIG. 11 is a flowchart showing a series of procedures for detecting a moving object having a known size.

FIG. 12 is a flowchart showing a series of procedures for detecting a moving body position.

FIG. 13 is a flowchart illustrating a series of procedures for detecting a distance to a moving object.

FIG. 14 is a flowchart showing a series of procedures of moving body size detection and moving body size correction.

FIG. 15 is a flowchart showing a series of procedures in a second embodiment showing details of the procedure of step S5 in FIG. 11;

FIG. 16 is a configuration diagram illustrating an example of a relationship between a size of the same object in an image and a distance from a camera.

17 is a configuration diagram showing the size of an object in a video in each state of FIG.

FIG. 18 is a configuration diagram illustrating a relationship between an imaging magnification and an angle of view.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Zooming lens 2 Focusing lens 3 Image sensor 4 Camera process unit 5 Moving object detection unit 6 Moving object size detection unit 7 Focus detection target area setting unit 8 Focus control unit 9 Zoom setting unit 10 Zoom control unit 11 Distance measurement unit 20 Overall Control unit 21 Moving object size correction unit 22 CPU 23 ROM 24 RAM 25 to 29 I / O 39 Interface with external device

Claims (19)

[Claims] 1. A moving object detecting means for detecting a moving object from image data input by an imaging means, a moving object size detecting means for detecting information on a size of a moving object detected by the moving object detecting means, and the moving object detecting means Measuring means for measuring the distance from the moving body to the imaging means detected in the above, a predetermined size from the distance information measured by the measuring means and information on the size of the moving body detected by the moving body size detecting means A moving body detecting device for detecting a moving body having the moving body. 2. The moving object detecting apparatus according to claim 1, wherein said predetermined moving object detecting means has setting means for setting information relating to the size of the moving object to be detected, and detects the moving object having the size set by said setting means. The moving object detection device according to any one of the preceding claims. 3. The predetermined moving object detecting means uses the measured distance information to obtain at least one of the information on the detected size of the moving object and the set information on the size of the moving object to be detected. 3. A moving object detection apparatus according to claim 2, further comprising a correcting means for correcting one of the moving objects, and detecting the moving object having the predetermined size using information on the size of the moving object corrected by the correcting means. apparatus. 4. A zoom control means for the image pickup means,
The correction means uses the magnification information and the measured distance information used in the zoom control means, and information on the size of the detected moving object and information on the set size of the moving object to be detected. The moving object detection device according to claim 3, wherein at least one of the following is corrected. 5. The apparatus according to claim 1, further comprising a focusing control unit provided in the imaging unit, wherein the measuring unit measures distance information using zoom control information in the zoom control unit and focusing control information in the focusing unit. The moving object detection device according to claim 4, wherein: 6. The zoom control information is information indicating a driving state of a zooming lens of the imaging unit, and the focusing control information is information indicating a driving state of a focusing lens of the imaging unit. The moving object detection device according to claim 5, wherein 7. The method according to claim 1, wherein the correction means corrects information on the size of the moving object to a size detected at a reference distance and at a reference magnification of the imaging means according to the measured distance information. The moving object detection device according to claim 3. 8. The moving object detection device according to claim 1, wherein the information on the size of the moving object is size information of the moving object. 9. The moving object detection apparatus according to claim 1, wherein the information on the size of the moving object is a region of a predetermined shape surrounding the moving object in the video. 10. A step of detecting a moving object from image data input by an imaging unit, a step of detecting information on the size of the detected moving object, and measuring a distance from the detected moving object to the imaging unit. And a step of detecting a moving object having a predetermined size from the measured distance information and the information on the detected size of the moving object. 11. The method according to claim 11, wherein the step of detecting a moving object having a predetermined size includes a step of setting information on the size of the moving object to be detected, and detecting the moving object having the set size. Item 11. The moving object detection method according to Item 10. 12. The step of detecting a moving object having a predetermined size includes using the measured distance information to obtain information on the size of the detected moving object and the set size of the moving object to be detected. 12. A moving object detection method according to claim 11, further comprising the step of correcting at least one of the information about the moving object and detecting the moving object having the predetermined size using the corrected information on the size of the moving object. Method. 13. The method according to claim 13, further comprising the step of controlling the zoom of the image pickup means, wherein the step of correcting includes determining the size of the detected moving object by using magnification information used in the zoom control and the measured distance information. 13. The moving object detection method according to claim 12, wherein at least one of the information on the moving object and the set information on the size of the moving object to be detected is corrected. 14. The method according to claim 1, further comprising the step of controlling focusing of the image pickup unit, wherein the measuring includes measuring distance information using zoom control information in the zoom control and focusing control information in the focusing control. 2. The method according to claim 1, wherein
4. The moving object detection method according to 3. 15. The zoom control information is information indicating a driving state of a zooming lens of the imaging unit, and the focusing control information is information indicating a driving state of a focusing lens of the imaging unit. The moving object detection method according to claim 14, wherein: 16. The method according to claim 1, wherein the correcting step corrects the information on the size of the moving object to a size detected at a reference distance and at a reference magnification of the imaging unit in accordance with the measured distance information. The moving object detection method according to claim 12, wherein 17. The information on the size of the moving object is as follows:
2. The information of the size of the moving object.
0 The moving object detection method according to 0. 18. The information on the size of the moving object is as follows:
The moving object detection method according to claim 10, wherein the moving image is a region of a predetermined shape surrounding the moving object in the video. 19. A computer-readable storage medium storing a program for executing the steps constituting the moving object detection method according to claim 10.