JPH10313455A - Supervisory image recording device and its reproducing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To always obtain appropriate image information and also to reduce records which have low application values by recording encoded image information and shape information. SOLUTION: An image is inputted from an input terminal 11; an object shape extracting means 12 detects whether an object is reflected by using an input image; when it is determined that an object exists, the shape of an detected object in the input image is extracted; and a shape image that represents an extracted shape is generated and outputted. An image encoding means 13 encodes an input image from the terminal 11. A shape encoding means 14 encodes shape information that is represented by a shape image 110 from the means 12 and outputs shape encoded data. The shape encoded data is sent to a 2nd input of a multiplexing means 15, and the means 15 multiplexes the shape encoded data and image encoded data from the means 13 and outputs them and stores them in a storage device 16.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surveillance image recording apparatus for photographing a surveillance area with a television camera and recording the photographed image, and a reproducing apparatus therefor.

[0002]

2. Description of the Related Art FIG. 11 is a block diagram showing a conventional surveillance image recording apparatus disclosed in Japanese Patent Application Laid-Open No. 9-27952. Specifically, an image is analyzed based on an image obtained by photographing with a television camera. 1 shows a monitoring image recording device that detects occurrence of an abnormality and records a compressed image in a recording device. In the figure, reference numeral 51 denotes an A / D converter for analog-to-digital conversion of an NTSC signal, which is an image captured by a television camera, and reference numeral 52 denotes a color space converter, which performs conversion while referring to a conversion table 53. 54 is a frequency space conversion unit for converting the color signal from the color space conversion unit 52 into a frequency conversion coefficient, and 55 is a quantization unit, which receives the frequency conversion coefficient from the frequency space conversion unit 54 and receives a quantization table 56 Is performed with reference to. Quantization unit 5
5 is sent to the inverse quantization unit 57, and at the same time,
The data is also sent to the encoding unit 58, subjected to encoding, and recorded in the storage device 59. The inverse quantization unit 57 performs inverse quantization by commonly referring to the quantization table 56 referred to by the quantization unit 55. Reference numeral 60 denotes a frequency space inverse transform unit that performs an operation reverse to that of the frequency space transform unit 54, and 61 denotes a color space inverse transform unit that performs an inverse operation of the color space transform unit 52.
Conversion is performed with reference to 2. Reference numeral 63 denotes a D / A converter for digital-to-analog conversion of a signal obtained by the color space inverse converter, which is transmitted to an external image receiving device such as a CRT.

Reference numeral 64 denotes a motion detecting unit for detecting a moving object, that is, a moving portion in the image from the color space inverse conversion unit 61; 65, a feature point extracting unit for detecting a portion satisfying a predetermined feature point in the image; Is an abnormality determination unit that determines an abnormality based on outputs from the motion detection unit 64 and the feature point extraction unit 65, outputs a warning signal, and outputs a control signal to the image control unit 67. Based on the control signal from the abnormality determination unit 66, the image control unit 67
3 and 62, control to change the contents of the quantization table 56.

Next, the operation will be described. The NTSC signal, which is an image captured by the TV camera, is converted into an A / D converter 5
At 1, the signal is converted from an analog signal to a digital signal, and sent to the color space converter 52. The NTSC signal has three components of a luminance signal and two kinds of color difference signals. The color space conversion unit 52 expresses the digitized NTSC signal, for example, the density of each component of the RGB color system for each pixel. To a color signal Pm (m = 0, 1,...). Note that the conversion by the color space conversion unit 52 is performed by a conversion table 5 in which conversion coefficients and the like are stored in a storage medium such as a ROM.
Conversion is performed with reference to FIG.

The color signal Pm output from the color space conversion unit 52 is transmitted to a frequency space conversion unit 54, and the frequency space conversion unit 54 performs, for example, a two-dimensional discrete cosine transform (DC).
T), and converts the color signal Pm, which is a component in the color space, into a frequency conversion coefficient Sij (i, j) which is a component in the frequency space.
= 0, 1,...). Frequency conversion coefficient Sij
(I, j = 0, 1,...) Are input to the quantization unit 55, and the quantization unit 55 quantizes the frequency transform coefficients. That is, each time the position of each frequency conversion coefficient Sij is specified (i, j), quantization is generally performed with a different step width (step size), and the step width for each spatial frequency is stored in a storage medium such as a ROM. The quantization is performed with reference to the quantization table 56 that is present. Here, in the quantization table 56, a coefficient Qij that defines a step width for each position (i, j) of the frequency conversion coefficient Sij is prepared. For example, the quantization unit 55 divides Sij by the coefficient Qij, and The quantization coefficient Rij is obtained by converting the quotient to an integer. The image quality and the amount of information can be adjusted by changing the value of the coefficient Qij, and the obtained quantization coefficient Rij (i, j = 0, 1,...)
5 is output.

[0006] The quantization coefficient Rij output from the quantization unit 55 is sent to the inverse quantization unit 57 and also to the encoding unit 58. The encoding unit 58 performs encoding such as variable-length encoding on the quantization coefficient Rij, outputs encoded data Hk, sends it to the storage device 59, and stores it.
The inverse quantization unit 57 performs an operation reverse to that of the quantization unit 55 to obtain the frequency transform coefficient S′ij, and executes the quantization by referring to the quantization table 56 referred to by the quantization unit 55 in common. Is done.

The frequency transform coefficient S′ij is calculated by a frequency space inverse transform unit 60 that performs the inverse operation of the frequency space transform unit 54.
And the color signal P′m is output. The color signal P'm is input to a color space inverse conversion unit 61 that performs an operation opposite to that of the color space conversion unit 52, and a digital NTSC signal is reconstructed. The color space inverse conversion unit 61 is, for example, a conversion table 6 in which conversion coefficients and the like are stored in a storage medium such as a ROM.
2 is performed with reference to FIG.

The digital NTSC signal (reconstructed image) output from the color space inverse converter 61 is converted from digital to analog by the D / A converter 63 and transmitted to an external image receiving device such as a CRT. The output from the color space inverse conversion unit 61 is also input to the motion detection unit 64 and the feature point extraction unit 65. The motion detection unit 64 detects a moving object, that is, a moving part in the image by performing an arithmetic process on the input reconstructed image. For example, by comparing the images between different frames, A moving object is detected. The feature point extracting unit 65 detects a portion that satisfies a predetermined feature point specified in advance in the image. For example, detection of the edge of the object is performed.

The detection results obtained by the motion detection section 64 and the feature point extraction section 65 are input to an abnormality determination section 66.
The abnormality determination unit 66 determines whether a predetermined condition corresponding to the occurrence of abnormality is satisfied based on the moving object detected by the motion detection unit 64 and the feature points detected by the feature point extraction unit 65. If the condition is satisfied, an alarm signal indicating the occurrence of an abnormality is sent to the outside, and a control signal for instructing the image quality control unit 67 to change the image quality is output. Image quality control section 67
Then, based on the control signal from the abnormality determination unit 66, the contents of the conversion table 53 quoted by the color space conversion unit 52, the contents of the quantization table 56 quoted by the quantization unit 56 and the inverse quantization unit 57, and further, The contents of the conversion table 62 referred to by the color space inverse conversion unit 61 are changed. For example, the conversion table 53, the quantization table 56, and the conversion table 62 store a plurality of types of contents in advance, and the image quality control unit 67 instructs the color space conversion unit 52 or the like to select which of them. I do.

Therefore, by changing the contents of each table cited as described above, the amount of image information output from the quantization unit 55 and encoded by the encoding unit 58 is changed. The image quality of the reconstructed image is changed. That is, the image quality control unit 67 controls each table so as to obtain a low-quality image with a high compression ratio in the normal period until the abnormality determination unit 66 detects the abnormality, and performs the compression in the abnormal period after the abnormality is detected. Control is performed to obtain a high-quality image with a low rate. Note that an external control signal from the outside is also input to the image quality control unit 67, and the operation is returned from the abnormal period to the normal period by instructing reset (cancellation) by an operation of a supervisor or the like. .

FIG. 12 is a diagram for explaining the operation of the conventional monitoring image recording apparatus. A rectangular frame schematically shows one frame of image information. This represents a transmission period (and a recording period) of the image information. In FIG. 12B, the high-quality image information HP
In FIG. 12A, low-quality image information LP having a small amount of information is obtained in a normal period, and high-quality image information HP having a large amount of information suitable for visual monitoring during an abnormal period in which an abnormality is detected. Is obtained.

[0012]

The conventional monitoring image recording apparatus records high-quality image information having a large amount of information only when the abnormality is determined to be abnormal by the abnormality determining section. It is configured to record a small amount of low-quality image information, and there is a problem that a high-quality image suitable for visual monitoring cannot be obtained unless the abnormality determination unit determines that there is an abnormality. Regardless of whether the image information is recorded or not, the image information is recorded, and an image of a low utility value in which no abnormality is detected is also recorded, and there is a problem that it is unsuitable for long-time recording.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems. An object is detected by using an input image to determine whether or not an object is present. By extracting the shape of the object and recording the image information and the shape information only within the shape of the object indicated by the extracted shape, it is possible to always obtain appropriate image information, and recording with low utility value It is an object of the present invention to obtain a surveillance image recording device and a reproduction device thereof that can reduce the number of images.

[0014]

A surveillance image recording apparatus according to the present invention detects whether or not an object is reflected by using an input image, extracts the shape of the detected object, and extracts an image representing the shape. Hereinafter referred to as a shape image), an object shape extracting means for generating an image in a shape in an input image based on the shape image, and a means for coding the shape image. It records encoded image information and shape information.

In the surveillance image recording apparatus according to the present invention, the object shape extracting means may detect the presence or absence of an object using an input image, based on an output of the object detecting means. Means for generating a shape image of the detected object.

In the surveillance image recording apparatus according to the present invention, the object detecting means may include means for storing a background image,
Means for obtaining an absolute value of a difference between the background image and the input image as a difference image; means for binarizing the difference image to divide the image into a change area and a still area; In the above case, it is determined that an object exists, otherwise, it is determined that no object exists, and a shape image representing the shape of the object is generated based on the shape information of the change area.

In the surveillance image recording apparatus according to the present invention, the object detecting means includes means for detecting whether or not the object is reflected by using the input image, and the encoding is performed only when it is determined that the object is present. It controls to record the image information and the shape information.

Further, in the reproducing apparatus according to the present invention, the reproducing apparatus for reproducing and decoding the recorded coded information comprises: means for decoding the encoded shape information; and decoding for decoding the encoded image information. Means for generating a decoded image by combining the decoded image information with a predetermined image based on the decoded shape information.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be specifically described with reference to the drawings showing the embodiments. Embodiment 1 FIG. FIG. 1 is a block diagram showing a monitoring image recording apparatus according to Embodiment 1 of the present invention. In the figure,
Reference numeral 11 denotes an input terminal to which an image captured by the television camera is input. The image input from the input terminal 11 is input to the object shape extracting means 12 and the image encoding means 13.
The output of the object shape extracting means 12 is given to the second input of the image coding means 13 and the input of the shape coding means 14. The multiplexing unit 15 receives the output of the image encoding unit 13 and the output of the shape encoding unit 14, multiplexes the encoded data, and supplies the multiplexed data to the input of the recording device 16.

FIG. 2 is a block diagram showing an example of the configuration of the object shape extracting means 12 in FIG. In the figure,
The input image is input to the sub-sampling unit 21, and the output of the sub-sampling unit 21 is provided to an input of the background image updating / storage unit 22 and a first input of the subtracter 23. Subtractor 2
The output of the background image updating / storing means 22 is given to the second input of No. 3. The output of the subtractor 23 is an absolute value calculator 24
Is input to The output of the absolute value calculator 24 is input to the binarizing means 25, the output of the binarizing means 25 is input to the comparator 27 via the area calculating means 26, and the output of the binarizing means 25 Is also supplied to the first input of the shape image generating means 28. The output of the comparator 27 is the shape image generating means 2
8, and the shape image generating means 28
A shape image representing the shape of the detected object in the input image is generated and output from the output of the value conversion means 25 and the output of the comparator 27. Note that the sub-sampling means 21 and the background image update /
Storage means 22, subtractor 23, absolute value calculators 24 and 2
The value converting means 25, the area calculating means 26 and the comparator 27 constitute the object detecting means 20.

FIG. 3 shows a background image updating / storing means 2 in the object detecting means 20 according to the first embodiment of the present invention.
FIG. 2 is a block diagram showing one configuration example. In the figure, the image output from the sub-sampling means 21 is given to a first input of a background image updating means 31, and the background image updating means 3
The output of 1 is given to the input of the background image storage means 32.
A second input of the background image updating means 31 is provided with a first output of the background image storing means 32. Background image storage means 3
The second output of No. 2 is output to the subtracter 23 as the output of the background image update / storage means 22.

Next, the operation will be described. An image captured by a television camera or the like is input from an input terminal 11 and is input to an object shape extracting unit 12 and an image encoding unit 13. The object shape extracting means 12 detects whether or not the object is reflected by using the input image, and when it is determined that the object is present, extracts the shape of the detected object in the input image, and forms a shape image representing the extracted shape. Is generated and output. This operation will be described with reference to FIG.

In the object shape extracting means 12, the input image is input to the object detecting means 20. Object detection means 20
First, the input image is sent to the sub-sampling means 21, and pixels are thinned out. For example, if the input image is an image of 640 pixels × 480 lines, the sub-sampling unit 21
Divides the input image into blocks of 4 pixels x 4 lines,
Each block is replaced with its average value, and 160 pixels × 1
Create and output an image of 20 lines. By the sub-sampling means 21, even when the resolution of the image recorded by the recording device 16 is high, the calculation amount of the object detecting means 20 can be reduced.

The difference between the image output from the sub-sampling means 21 and the background image is obtained in a subtractor 23. The background image is stored in the background image updating / storing unit 22 and is sequentially updated according to the image output from the sub-sampling unit 21. The absolute value of the difference output from the subtractor 23 is obtained by an absolute value calculator 24. The image output from the absolute value calculator 24 is a difference absolute value between the subsampled input image and the background image, but for simplicity, this image is hereinafter referred to as a difference image.

The operation of the background image updating / storing means 22 will now be described with reference to FIG. The input image whose pixels have been thinned out by the sub-sampling unit 21 and the background image stored in the background image storage unit 32 are input to the background image updating unit 31. Here, assuming that the input image after the sub-sample at time t is I (t) and the background image is B (t), the background image updating means 31 sets d (t) = I (t) − Find B (t-1). d (t)
B (t) if the absolute value of is less than or equal to a first predetermined value θ
= I (t), otherwise a second predetermined value ε is used, and when d (t) <0, B (t) = B (t-1) −ε, d (t )
If ≧ 0, update the background image as B (t) = B (t−1) + ε,
The updated background image B (t) is output to the background image storage means 32. The background image storage means 32 stores the background image B (t) and outputs the background image B (t) to the subtracter 23. Note that the initial value B (0) of the background image stored in the background image storage unit 32 is
The monitoring image recording apparatus according to the present invention may be separately acquired when it is installed, or may be an image having all pixel values of 0, 128, or 255.

Next, the difference image output from the absolute value calculator 24 is binarized by a binarizing means 25 using a predetermined value α. For example, for each pixel of the difference image, if the pixel value is equal to or less than α, 0 is output as that pixel is not a change area, and if the pixel value is larger than α, the pixel is included in the change area. 1 is output to form a binarized image. Next, in the area calculating means 26, the area of the change region is obtained. As a method of calculating the area, the number of pixels having a pixel value of 1 in the image output from the binarizing unit 25 may be counted. This area is compared with a predetermined value β determined in advance by the comparator 27. If the area is larger than β, it is determined that the object is displayed. If the area is smaller than β, it is determined that the object is not displayed. Then, a control signal c is output and sent to the shape image generating means 28. Here, the predetermined value β may be determined according to the size of the object to be determined in the screen.

An image obtained by binarizing the difference image output from the binarization unit 25 is also input to a shape image generation unit 28, and the shape image generation unit 28 compares the control signal c from the comparator 27 with the difference signal. A shape image representing the shape of the detected object is generated based on the binarized image of the image. That is, when the control signal c from the comparator 27 is a signal that determines that an object is reflected, the pixel whose pixel value of the binarized image from the binarization unit 25 is 1 is an object. The pixel value of the pixel within the shape of “1” is, for example, “1”, and the pixel value of the binarized image having the pixel value of “0” is set to “0” as the pixel of the background portion outside the shape, and the shape image 110 Generate and output At this time, since the image in the binarizing means 25 is obtained from the image obtained by thinning out the input image by the sub-sampling means 21, the shape image generating means 28
The shape image is generated by up-sampling the binary image by interpolation.

FIG. 4 is a diagram showing an example of an image for explaining the operation of the object shape extracting means 12 in the surveillance image recording device according to the first embodiment of the present invention. An example in the case where an object corresponding to a person exists in a car is shown. In this case, whether or not an object such as a person is present in the car corresponds to the monitoring target,
Assuming that FIG. 4A is an input image and FIG. 4B is a background image, the absolute value calculator 24 obtains these difference images. Shape image 110 obtained from the binarized image
Is an image as shown in FIG. That is, FIG.
In (c), the shape of the person A is a black portion in the figure, the pixel value within the shape is, for example, “1”, the background portion outside the shape is a white portion, and the pixel value outside the shape is “0”. As a result, a shape image 110 is generated.

When the control signal from the comparator 27 is a signal that determines that an object is not reflected, a shape image 110 whose pixel values are all "0" is generated and output.

The output of the object shape extracting means 12 is output to an image coding means 13 and a shape coding means 14.
The image encoding means 13 encodes the input image from the input terminal 11. A shape image 110 which is the output of the object shape extraction means 12 is input. The pixels in the shape corresponding to the pixel position of the pixel value “1” of 110 are encoded.

FIG. 5 is a diagram showing an example for explaining an encoding method in the image encoding means 13 in the monitoring image recording apparatus according to the first embodiment of the present invention. For example, as shown in FIG. Is divided into blocks (encoded blocks) having a predetermined number of pixels and encoded in block units. Here, a portion p shown in black in the figure indicates a pixel at a position corresponding to the shape shown in the shape image 110, and the coding is performed when all the pixels in the block are pixels in the shape. DCT coding is performed on a block basis only for a block including a boundary pixel with the background “0”. At this time, for a block including the background “0” and the boundary pixel, pixels outside the shape, for example, the pixel value of the boundary pixel is repeatedly padded (padded) and then subjected to DCT coding. Therefore, the blocks to be coded are only the blocks in the shape p in FIG. 5 and the blocks indicated by oblique lines, and blocks in which all the pixels in the block are out of the shape are skipped during coding. Therefore, only the image information within the valid shape is output as encoded data. Then, the encoded image data output from the image encoding unit 13 is sent to a first input of the multiplexing unit 15.

On the other hand, the shape encoding means 14 encodes the shape information represented by the shape image 110 from the object shape extracting means 12 and outputs shape encoded data. In the shape encoding, for example, the shape image 110 is horizontally scanned to detect a change position of a pixel whose pixel value changes from “0” to “1” or “1” to “0”, and a position of each change point is detected. The relationship is encoded to form shape-encoded data. FIG. 6 is a diagram showing an example of this. When the shape image 110 is scanned and, for example, changed pixels a and b are detected, the number of pixels between a and b is encoded to form encoded shape data.

Then, the shape-encoded data output from the shape-encoding means 14 is sent to a second input of the multiplexing means 15, and the multiplexing means 15 outputs the shape-encoded data to the image encoding means 13 Are multiplexed and output, and are stored in the storage device 16. As the storage device 16, a hard disk, a magneto-optical disk, a magnetic tape,
TR or the like is used.

When the output of the object detecting means 20 in the object shape extracting means 12 indicates that no object is reflected, the shape image 1 whose pixel values are all "0" is displayed.
10 is output, at this time, all blocks are skipped by the image encoding means 13, and the shape encoding means 14 does not have any pixels indicating the shape in the shape image 110, so that the shape encoding is not performed. No data is output, and thus the storage device 16 does not record shape-encoded and image-encoded data.

Therefore, what is stored in the storage device 16 is the shape-encoded data and the image-encoded data when the object is detected by the object shape extracting means 12 and the shape image is extracted, and the object is detected. If the shape image does not indicate that there is a shape, no data is stored, and only image information within the shape of the detected object and the shape information are stored.

In the first embodiment, a case has been described in which the image encoding means 13 performs DCT encoding in units of blocks and performs padding and encoding of pixels outside the shape, but the image encoding means 13 However, the present invention is not limited to this, and encoding may be performed by an encoding method using another shape such as a wavelet transform or a shape adaptive DCT (SA-DCT) that performs transform according to pixels in the shape.

Further, in the first embodiment, the shape encoding means 14 encodes the positional relationship of the changing pixels whose pixel values change from outside the shape to inside the shape or from inside the shape to outside the shape in the shape image 110. It is said that encoding is performed by performing, but the encoding method of the shape image is not limited to this,
For example, a method may be used in which contour points of the shape of the object are sequentially encoded, or encoded at a relative position to a previous change point on an encoding line in the shape image by line correlation of the image.

In the first embodiment, the sub-sampling unit 21 replaces the input image with the average value for each block of 4 pixels × 4 lines.
Is not limited to this, and may be simple pixel thinning or pixel thinning using a low-pass filter. Further, the sub-sampling means 21 is for reducing the amount of calculation of the object detecting means 20. Therefore, if there is sufficient calculation time, the sub-sampling means 21 need not perform the thinning.

Further, in the first embodiment, the binarizing means 25 binarizes the difference image with a predetermined value α, but the predetermined value α is adaptively changed according to the input image. It may be configured to provide the same effects as above.

Further, in the first embodiment, the background image updating / storing means 22 in the object detecting means 3 is configured as shown in FIG. 3, but the updating and storing of the background image is not limited to this. To detect changes to the image
A configuration may be used in which the image in the frame memory is updated as a background when there is no change for a predetermined time, and otherwise the image is not updated.

In the first embodiment, the pixel value in the shape in the shape image 110 is set to the value 1 and the pixel value outside the shape is set to 0, but the present invention is not limited to this. Other pixel values may be used.

Further, the object shape is extracted by the object shape extracting means 12 by binarizing the difference image by the object detecting means 20 and calculating the area of the change area by the area calculating means 26 to detect the object. The shape image is generated from the binarized image and the detection result. However, the present invention is not limited to this. For example, a closed area is obtained by detecting an edge of the image, and a shape is extracted. Detects motion in the image, finds a motion area from the detected motion,
The same effect can be obtained if the shape of the object in the input image can be extracted and a shape image representing the shape can be generated, such as a method of obtaining a shape from this region.

Embodiment 2 FIG. 7 is a block diagram showing a surveillance image recording device according to a second embodiment of the present invention, and FIG. 8 is a block diagram showing a configuration example of an object shape extracting means 12b of the surveillance image recording device according to the second embodiment of the present invention. It is. In the first embodiment, the output of the object shape extracting means 12 is configured to be the detected shape image 110 of the object.
As shown in FIGS. 7 and 8, the object shape extracting means 12b
Is output as the control signal c from the shape image 110 and the comparator 27 in the object detection means 20. When the control signal c indicates that an object is reflected in the input image, the output from the multiplexing means 15 It is also possible to record the shape-encoded data and the image-encoded data, which are the outputs of (1) and (2), if the control signal c indicates that the object is not reflected, the recording is not performed.

In FIG. 7, reference numerals 11, 13 to 15 are the same as those of the monitoring image recording apparatus according to the first embodiment.
12b is an object shape extracting means, and 16b is a storage device.
In FIG. 8, reference numerals 20 to 28 are all the same as those in the object shape extracting means 12 in the first embodiment.

Next, the operation will be described. An image captured by a television camera or the like is input from an input terminal 11, and an object shape extraction unit 12b detects an object in the input image by an object detection unit 20 and controls a comparator 27 to indicate whether the object is reflected. An operation of outputting the signal c and generating and outputting the shape image 110 in the shape image generating means 28;
The encoding of the image by the image encoding unit 13, the encoding of the shape image by the shape encoding unit 14, and the operation of the multiplexing unit 15 are the same as those in the first embodiment, and therefore the detailed description is omitted. I do.

The comparator 2 in the object shape extracting means 12b
Control signal c indicating whether an object from 7 is reflected
Is also provided to a second input of the storage device 16b. When the control signal c from the object detection means 12b indicates that an object is reflected, the storage device 16b records the shape encoded data and the image encoded data output from the multiplexing means 15 and controls If the signal c indicates that no object is shown, no recording is performed.

Therefore, the data stored in the storage device 16b is a case where it is determined that an object is reflected in the input image, otherwise the storage operation is stopped, and the shape of the detected object is The encoded data of the image and the encoded data of the shape are stored.

Embodiment 3 Hereinafter, a third embodiment of the reproducing apparatus that reproduces and decodes the encoded data stored by the monitoring image recording apparatus according to the first and second embodiments will be described. FIG. 9 is a block diagram showing an example of the configuration of a playback device that plays back and decodes encoded data stored by the monitoring image recording device according to the first and second embodiments. In the figure, reference numeral 16 denotes a storage device on which encoded data similar to those of the first and second embodiments are recorded, and 41 denotes a separation device for separating encoded data read from the storage device 16 into shape encoded data and image encoded data. Means, and the separated shape-encoded data is provided to an input of the shape decoding means 42, and an output thereof is provided to a first input of the synthesizing means 44. The encoded image data is supplied to the input of the image decoding means 43, and the output is input to the second input of the synthesizing means 44. The synthesizing means 44 forms a decoded image and sends it to the output terminal 45.

Next, the operation will be described. Storage device 16
The coded data read from the data is input to the separating means 41 to separate the shape coded data s and the image coded data i, and the shape coded data s is sent to the shape decoding means 42 and the image coded data i Outputs respective data to the image decoding means 43. The shape decoding means 42 decodes the input shape encoded data s, generates and outputs an original shape image (decoded shape image) 111, and sends it to the synthesizing means 44.
The image decoding means 43 decodes the encoded image data and sends it to the synthesizing means 44. In the synthesizing unit 44, based on the decoded shape image 111 from the shape decoding unit 42, the decoded data from the image decoding unit 43 is obtained as a pixel value of an image corresponding to the shape shown in the decoded shape image 111, A pixel value outside the shape is set to a background image having a pixel value of 128, for example, and the background image and the image within the shape are combined to obtain a decoded image 11.
2 is generated and output to the output terminal 45.

For example, when the shape image 111 obtained by decoding by the shape decoding means 42 is as shown in FIG. 10A, the image encoded data is data in which the pixels in the shape are encoded. When the image is synthesized based on the decoded shape image 111, the image in the shape as shown in FIG.
Thus, a decoded image 112 synthesized with the background image 8 can be obtained.

In the third embodiment, the combining means 44
When the decoded image is generated by using the pixel value 128 as the background image
Has been described, but an image used as a pixel outside the shape may be an image having all pixel values of 0 or 255, etc.
Alternatively, an image that has been previously captured and stored may be used.

[0052]

Since the present invention is configured as described above, it has the following effects.

According to the surveillance image recording apparatus of the present invention, it is detected whether or not an object is reflected by using the input image, the shape of the detected object is extracted, and a shape image representing the shape is generated. Encoding an image in a shape in an input image based on the shape image, and encoding the shape image, and recording the encoded image information and shape information, thereby inputting an image from a television camera or the like. Only by recording the image information in the shape of the object detected from the input image, it is possible to always obtain the image information in the shape of the object while the object is present in the monitoring area. Recording of information can be reduced.

According to the surveillance image recording apparatus of the present invention, the object shape extracting means detects whether or not an object is reflected by using the input image, and based on the detection result,
By generating a shape image representing the shape of the detected object, the image information within the shape of the object detected from the input image is recorded, and the image information within the shape of the object is always stored while the object is present in the monitoring area. Records that have low utility value outside the object can be reduced.

According to the surveillance image recording apparatus of the present invention, the object detecting means stores the background image, determines the absolute value of the difference between the background image and the input image as a difference image, and calculates the difference image by two. The value is divided into a change area and a stationary area, and when the area of the change area is equal to or more than a predetermined value, it is determined that the object exists, otherwise, it is determined that the object does not exist, and the shape information of the change area is determined. By configuring to generate a shape image representing the shape of the object, the image information in the shape of the object detected from the input image is recorded, and the image in the shape of the object is always present while the object is present in the monitoring area. Information can be obtained, and records with low utility value other than inside the object can be reduced.

According to the surveillance image recording apparatus of the present invention, the object detecting means includes means for detecting whether or not the object is reflected using the input image, and the encoding is performed only when it is determined that the object is present. By controlling to record the image information and the shape information, only the image information within the shape of the object detected from the input image is recorded, and the image information within the shape of the object is always obtained while the object exists in the monitoring area. And records with low utility values other than inside the object can be reduced.

According to the reproducing apparatus for decoding the encoded information recorded by the monitoring image recording apparatus according to the present invention, the encoded shape information is decoded, and the encoded image information is decoded. Based on the decoded shape information, by combining the decoded image information with a predetermined image to generate a decoded image, only the image information within the shape of the object detected from the input image is recorded, and Image information within the shape of the object in the recorded input image can be obtained,
Records with low utility value other than inside the object can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a monitoring image recording device according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration example of an object shape extraction unit 12 in FIG.

FIG. 3 is a block diagram showing a configuration example of a background image update / storage unit 22 in the object detection unit 20 according to the first embodiment of the present invention.

FIG. 4 is a diagram showing an example of an image for explaining the operation of the object shape extracting means in the monitoring image recording device according to the first embodiment of the present invention.

FIG. 5 is a diagram illustrating an example for describing an encoding method of an image encoding unit in the monitoring image recording device according to the first embodiment of the present invention.

FIG. 6 is a diagram illustrating an example for describing an encoding method of a shape encoding unit in the monitoring image recording device according to the first embodiment of the present invention.

FIG. 7 is a block diagram showing a surveillance image recording device according to a second embodiment of the present invention.

FIG. 8 is a block diagram showing a configuration example of an object shape extracting means 12b of the monitoring image recording device according to the second embodiment of the present invention.

FIG. 9 is a block diagram showing a configuration example of a reproducing apparatus in a monitoring image recording apparatus according to a third embodiment of the present invention.

FIG. 10 is a diagram showing an example of a decoded shape image and a decoded image for explaining the operation of the reproducing device in the monitoring image recording device according to the third embodiment of the present invention.

FIG. 11 is a block diagram showing a conventional monitoring image recording device.

FIG. 12 is a diagram for explaining the operation of a conventional monitoring image recording device.

[Explanation of symbols]

11 input terminal, 12, 12b object shape extracting means, 1
3 image encoding means, 14 shape encoding means, 15 multiplexing means, 16, 16b storage device, 20 object detection means, 21 sub-sampling means, 22 background image updating / storage means, 23 subtractor, 24 absolute value calculator , 25 binarizing means, 26 area calculating means, 27 comparators, 28 shape image generating means, 41 separating means, 42 shape decoding means, 43 image decoding means, 44 combining means, 45 output terminals, 110 shape images, 111 decoded shape image, 1
12 Decoded image.

Claims (5)

[Claims] 1. An object shape extracting means for detecting whether an object is projected using an input image, extracting a shape of the detected object, and generating an image representing the shape, Means for encoding an image within the shape in the input image based on the image representing the shape, and means for encoding an image representing the shape of the detected object, wherein the encoded image information and shape information are A surveillance image recording device, characterized by recording an image. 2. An object shape extracting means for detecting whether or not an object is projected using an input image, and an image representing the shape of the detected object based on an output of the object detecting means. The surveillance image recording device according to claim 1, further comprising: a generation unit. 3. The object detection means includes means for storing a background image, means for obtaining an absolute value of a difference between the background image and the input image as a difference image, and a method for binarizing and changing the difference image. Means for dividing into an area and a stationary area, wherein when the area of the change area is equal to or larger than a predetermined value, it is determined that an object is present, and an image representing the shape of the object is generated based on the shape information of the change area. 3. The monitoring image recording apparatus according to claim 2, wherein: 4. The apparatus according to claim 1, wherein said object detecting means includes means for detecting whether or not the object is reflected by using an input image, and records encoded image information and shape information only when it is determined that the object is present. The surveillance image recording apparatus according to claim 2, wherein the control is performed. 5. A reproducing device for decoding encoded information recorded by the monitoring image recording device according to claim 1, wherein the reproducing device decodes the encoded shape information. Decoding means for decoding the encoded image information; and means for generating a decoded image by combining the decoded image information with a predetermined image based on the decoded shape information. Playback device.