JP3393168B2 - Image input device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image input device suitable for use in an image transmission device such as a TV (television) conference device and a TV telephone device.

[0002]

2. Description of the Related Art Conventionally, a CCD (charge coupled device) has been used in a computer system such as a personal computer or a workstation, a home information terminal such as a so-called home automation system, or a portable information terminal.
A TV camera including the above as an image pickup device is connected as an image input means, and an image picked up by the TV camera is transmitted to transmit the image.
A configuration adapted to realize a V-conference device and a TV phone device has become widespread.

In a TV camera used for such an application, the subject that needs to be imaged is a person, especially the face and materials such as documents and drawings. Therefore, it is desired to magnify a specific subject in order to pay attention to these subjects and recognize them in detail. Therefore, the above TV
The camera has an optical zoom mechanism and the like.

However, if the subject to be magnified is not in the center of the screen, there is a problem that the desired specific subject cannot be appropriately magnified by simply magnifying using the zoom mechanism. .

Conventional techniques for solving such problems are disclosed in JP-A-6-202600 and JP-A-7-200600.
In the one disclosed in Japanese Patent No. 135594, the optical zoom mechanism is provided in the TV camera, and the platform is a movable platform capable of pan / tilt operation.

[0006]

Therefore, in the above-mentioned conventional technique, since the platform is movable, a motor and its control circuit are required, which results in high cost and large size, and complicated operation. There is a problem.

It is an object of the present invention to provide an image input device which can be reduced in cost and size, and which can extract and enlarge a desired specific image by a simple operation.

[0008]

In order to achieve the above-mentioned object, the image input device of the present invention has a first aspect of the invention, and an image input means for picking up an image of a subject and outputting the image, and the image input means. Image dividing means for dividing the image from the image into predetermined divided images, and image designation for specifying a single or a plurality of divided images including a part of a specific subject from among the divided images divided by the image dividing means. Means, and among the divided images divided by the image dividing means, a single or a plurality of divided images that include the specified divided image specified by the image specifying means and include the entire specific subject are specified. It is characterized in that it includes image specifying means for recognizing as divided images.

According to the above-mentioned structure, in the image input device used for the TV conference device or the like, from the image of the subject imaged by the image input means such as the TV camera or the like, the person by the image designating means, especially the face or the document thereof, etc. By specifying the divided image including a part of the above, the image specifying unit recognizes the divided image including the person, particularly the whole face or document as the specified image.

According to the invention of claim 2, an image input means for picking up an image of a subject and outputting the image, an image dividing means for dividing the image from the image input means into predetermined divided images, and the image division. From the divided images divided by the means, an image designating means for designating a single or a plurality of divided images including a part of a specific subject, and the image selected from the divided images divided by the image dividing means. An image specifying unit that recognizes a single or a plurality of divided images that include the specified divided image that is specified by the specifying unit and that includes the entire specific subject as the specific divided image, and that is recognized by the image specifying unit. An image selecting unit capable of selecting a specific divided image is included.

According to the above construction, in addition to the operation of the construction described in claim 1, a plurality of divided images including a person specified by the image specifying means by the image selecting means, particularly the entire face or document thereof, can be obtained. You can choose from patterns. Therefore, for example, when a part of the upper half of the person is designated, it is possible to arbitrarily instruct to widen or narrow the specific range such as the face, the upper half of the body, or the whole body.

According to the third aspect of the present invention, image input means for picking up an image of a subject and outputting the image, image division means for dividing the image from the image input means into predetermined divided images, and the image division. From the divided images divided by the means, an image designating means for designating a single or a plurality of divided images including a part of a specific subject, and the image selected from the divided images divided by the image dividing means. An image specifying unit that recognizes a single or a plurality of divided images that include the specified divided image that is specified by the specifying unit and that includes the entire specific subject as the specific divided image, and that is recognized by the image specifying unit. Image selecting means capable of selecting a specific divided image, and focus and iris of the image input means based on image data in the selected specific divided image selected by the image selecting means. It is characterized in that and a control means for controlling the white balance and the like.

According to the above construction, in addition to the operation of the second construction, a single pattern including the person selected by the image selection means by the camera control means, in particular, the entire face, document, etc. Based on the brightness data and color temperature data in the divided image, focus, iris,
The white balance etc. can be controlled. Therefore, for example, with respect to the focus, even if the specific subject to be watched is not in the center of the entire image, the focus is controlled so as to focus on the selected divided image, so that a clear image can be obtained.

The image input device according to the invention of claim 4 further comprises image input means for picking up an image of a subject and outputting the image, and image division for dividing the image from the image input means into predetermined divided images. Means, an image designating means for designating a single or a plurality of split images including a part of a specific subject from the split images split by the image splitting means, and a split image split by the image splitting means. Image specifying means for recognizing, as a specific divided image, a single or a plurality of divided images including the specified divided image specified by the image specifying means, and including the entire specified subject, An image selecting means capable of selecting the specific divided image recognized by the specifying means, and the image input means based on the image data in the selected specific divided image selected by the image selecting means. Focus, iris and is characterized in that it comprises a control means for controlling the white balance and the like, and image processing means for expanding the selected specific divided image selected by the image selection means.

According to the above construction, in addition to the operation of the construction according to the third aspect, a single pattern including the person selected by the image selection means by the image processing means, in particular, the entire face, document, etc. Only the split image can be enlarged. Therefore, it is possible to easily recognize the details of a person whose controlled focus, iris, white balance and the like are in the best state, especially the divided image of a single pattern including the entire face and documents.

Further, the image input device according to the invention of claim 5 is an image input means for picking up an image of a subject and outputting the image, and an image dividing means for dividing the image from the image input means into predetermined divided images. An image designating unit that designates a single or a plurality of segmented images including a part of a specific subject from among the segmented images segmented by the image segmenting unit, and a segmented image segmented by the image segmenting unit. From among these, an image specifying unit that recognizes a single or a plurality of divided images that include the specified divided image specified by the image specifying unit and that includes the entire specific subject, and the image specifying unit. The image selecting means capable of selecting the specific divided image recognized by the means, and the image input means based on the image data in the selected specific divided image selected by the image selecting means. Carcass, iris, characterized by comprising control means for controlling the white balance and the like, and an image storage means for storing the selected specific divided image selected by the image selection means.

According to the above construction, in addition to the operation of the construction according to the fourth aspect, a single pattern including the entire person selected by the image selection processing by the image storage processing means, in particular, the face, documents, etc. Only the divided images of can be stored. Therefore, it is possible to easily and precisely recognize the details of a person whose focus, iris, white balance, and the like are controlled in the best state, in particular, a single-pattern divided image including the entire face and documents as a still image.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of an image input apparatus of the present invention will be described below with reference to the drawings. First, a first embodiment of the image input device of the present invention will be described with reference to FIGS.

FIG. 1 is a block diagram showing an electrical configuration of a first embodiment of an image input device of the present invention. The image input device 10 is used as a camera of a TV conference device, for example. In general, the camera unit 12, the image dividing unit 13, the image designating unit 14, and the image specifying unit 15
And an image selection unit 16 and a camera control unit 17.

The camera unit 12 which is an image input means is
It is realized by a D digital television camera or the like, images a subject 11 such as a person or a document such as a document or a drawing, and outputs image data corresponding to the captured image to the image dividing unit 13.

The image division unit 13 divides the image picked up by the camera unit 12 into a large number of areas as shown in FIG. 2, for example. In the example of FIG. 2, (6) of A1, A2, ..., A48.
The example shows division into (8) = 48) regions. Further, FIG. 3 shows an example in which the captured image is divided into the areas A1 to A48 (hereinafter referred to as reference numeral A when collectively referred to).

The image designating section 14 designates, to the image specifying section 15, a single or a plurality of adjacent areas of each area A in the image dividing section 13. Using a not-shown image display device that displays a captured image without displaying division boundaries and a not-shown bonding device such as a touch panel or a mouse, operate a single area or a plurality of adjacent areas from each area A. The person specifies the divided area without being aware of it.

The image specifying unit 15 analyzes the image from the image dividing unit 13 and includes a single or a plurality of regions designated by the image designating unit 14 among the divided images, and also includes a specific image. The divided image to be recognized is recognized as a specific image, and the specified partial image data is output.

For example, in the example of the imaged image shown in FIG. 3, when the face P1 of a person is used as the specific image, areas A10, A11, A12; A18, A1 forming the face P1 are formed.
When the divided images of 9, A20; A26, A27, and A28 are recognized as the specific image and the image P2 of the person is set as the specific image, areas A10, A11, A12; A18, A1
9, A20; A26, A27, A28; A34, A3
5, A36; A42, A43, A44 divided images are recognized as specific images.

When the document P3 is used as a specific image, areas A22, A23, A24; A30, A31, A
32; when the divided images of A38, A39, and A40 are recognized as the specific image and the light source P4 is set as the specific image, A
5, A6; A13, A14 divided images are recognized as specific images.

The image selecting section 16 receives the specified partial image data output from the image specifying section 15, stores a plurality of types of the image data, and uses a touch panel or a pointing device such as a mouse (not shown). ,
The operator can select one type from a plurality of types of image data and outputs the selected partial image data.

The camera control unit 17 has the image selection unit 16 described above.
The focus, iris, white balance, etc. of the camera section 12 are controlled by outputting a control signal based on a luminance signal, a color signal, etc. of the partial image data selected in.

FIG. 4 is a flow chart for explaining the operation of the image specifying unit 15. The operation will be described below.
In step S1, the designated divided area designated by the image designating unit 14 is set as an image in the edge extraction area. In step S2, the edge of the subject is extracted from the change in the brightness level in the edge extraction area.

Then, in step S3, Japanese Patent Laid-Open No. 7-6
By applying the method described in No. 5149 or the like, it is determined whether or not the divided image representing the face P1 exists in this edge extraction area. That is, first, an absolute difference between image data separated by a predetermined frame is obtained to detect a moving area.

Next, the data of the difference value is density projected onto the X-axis coordinate (horizontal direction) and the Y-axis coordinate (vertical direction), and the top of the moving region (maximum value in the Y-axis direction) is topped. The coordinates are used, and the X coordinate from which the parietal coordinates are obtained is the center coordinate of the human body, and the edge of the difference value within the predetermined range in the X coordinate from the center coordinate is the human face candidate position.

When the face P1 is present, the divided image forming the face P1 is extracted as a specific image in step S4 (face extraction area determination), and in step S5, the data of the extracted divided image is extracted. Is output.

Further, when the face P1 cannot be specified in step S3, the process proceeds to step S6, and it is determined whether or not there is an image to be specified other than the face P1 such as the document P3. For example, when the image to be specified is the entire object, it is determined whether or not a closed region exists at the extracted edge. When there is an image to be specified, after the divided images forming the specified image are extracted in step S7 (object extraction area confirmation), the data of the extracted divided images is output in step S5.

On the other hand, in step S6, when the image to be specified is not detected, that is, when the specified divided area specified by the image specifying unit 14 includes only a part of the specified subject, Go to step S8. In step S8, the designated divided area is enlarged and set as the extraction area, and the process returns to step S2. The image recognition is performed in this way, and the image specifying unit 15 outputs the image data of the recognized image of the face, the document, etc. as the specific image as described above.

FIG. 5 is a recognition example for explaining the operation of the image specifying unit 15. In this example, the image designation unit 1 of FIG.
When the operator specifies area 10 (A10) in 4,
It shows how the face P1 is recognized. First, in N1, the image is specified only in the area 10 (A10), but the face P1 cannot be detected.

Therefore, the recognition range is expanded to A1 in N2.
The image is specified in four areas of 0, A11; A18, A19, but the face P1 is not detected yet. Further, the recognition range is expanded, and in N3, A10, A11, A12; A1
When the image is specified in the nine areas A8, A19, A20; A26, A27, A28, the face P1 is detected and the recognition ends. The specified partial image data of the nine areas is output from the image specifying unit 15 in FIG.

The method of expanding the recognition range in this example is merely an example for explanation, and the range may be expanded centered around the designated area. Further, although the case where the designated area is single is described, the same applies to the case where a plurality of adjacent areas are provided. Further, when the designated area itself is plural, it can be realized by repeating the above recognition procedure plural times.

FIG. 6 is an operation example for explaining the operation of the image selecting section 16. In this example, the image designation unit 1 of FIG.
When the operator specifies area 10 (A10) in 4,
It shows how the face P1 and the person P2 can be selected. The process up to the end of recognition A from N1 to N3 in the upper part is the same as the recognition example of FIG.

At N4 in the lower stage, recognition is started again after recognition A ends, and the recognition range is expanded to A10, A11, A1.
2; A18, A19, A20; A26, A27, A2
8; When the image is specified in the 12 areas A34, A35, and A36, the person P2 is not detected yet. Further, the recognition range is expanded and in N6, A10, A11, A12; A.
18, A19, A20; A26, A27, A28; A3
When the images are specified in 16 areas A4, A35, A36; A42, A43, A44, the person P2 is detected and the recognition B ends.

As described above, the operator can select whether the face P1 is the specific image or the person P2 is the specific image by selecting the result of the recognition A and the result of the recognition B. After the selection, the partial image data selected by the image selection unit 16 in FIG. 1 is output.

As in the case of FIG. 5, the method of expanding the recognition range in this example is merely an example for explanation, and the range may be expanded centered around the designated area. Further, although the case where the designated area is single is described, the same applies to the case where a plurality of adjacent areas are provided. Further, when the designated area itself is plural, it can be realized by repeating the above recognition procedure plural times.

When the image input device 10 according to the present invention is applied to the camera of the TV conference device in this way, when transmitting image data from the transmission circuit to the display device or the like via the communication line and the reception circuit, the communication circuit, etc. Since the specific image data is output for the data of all the images defined by the transmission capacity of, the transmission capacity can be effectively used.

When the number of pixels in each area A is small, that is, when the original image is divided into a large number of divided images, unnecessary portions such as the background other than the specific subject can be reduced.

Next, a second embodiment of the image input apparatus of the present invention will be described with reference to FIGS. Figure 7
6 is a block diagram showing an electrical configuration of an image input device 20 according to a second embodiment of the present invention.
0 is similar to the image input device 10 described above, and corresponding parts are designated by the same reference numerals and description thereof is omitted.

It should be noted that in the image input device 20, the specified partial image data and the selected partial image data from the image specifying unit 15 and the image selecting unit 16 are displayed in the image enlarging unit 28, for example, as shown in FIG. That is, an enlarged image of the partial image is obtained as shown in FIG.

When the area 10 (A10) is designated (when the face is selected) with respect to the original image (original image) of FIG. 8A, the face selection shown in FIG. 8B is enlarged. When an image is obtained and the area 42 (A42) is instructed or the area 10 (A10) is instructed (a person is selected), an enlarged image of person selection shown in FIG. 8C is obtained, When the area 22 (A22) is instructed, FIG.
An enlarged image shown in (d) is obtained. In this example, the image is enlarged to the size of the entire original image, but the enlargement factor may be predetermined or the operator may arbitrarily select it.

The image enlarging section 28 is, for example, disclosed in Japanese Patent Laid-Open No. 7-
It is configured as described in Japanese Patent No. 288781. That is, the image data before conversion is alternately written into the two line memories line by line, and the output is
Based on the correspondence between the line before conversion and the line after conversion, the output from only one of the line memories or the output from the two line memories is weighted.

For example, when the enlargement ratio is 2.5, it is necessary to convert the image data of two lines into five lines, and the first and second lines have the first line. The image data from the line memory only is output, and the image data from the first and second line memories are added with a weight of 1/2 and output to the third line, and the fourth and fifth lines are output. Image data from only the second line memory is output to the line. Similarly, the image data of each dot in the horizontal direction is also converted in correspondence with the number of dots before conversion and the number of dots after conversion. Note that the image enlarging method is not limited to such a method, and is, for example, Japanese Patent Laid-Open No. 4-1
Other methods such as those described in Japanese Patent No. 85087 may be used.

The image storage unit 29 is a still image memory for storing the image data enlarged by the image enlarging unit 28, and the position of the original image is changed in real time in the partial image data which is once specified or selected. However, the original image can be reproduced.

In this way, the image input device 20 according to the present invention extracts a specific image of a person or the like from the original image as shown in FIG. 3, and as shown in FIG. Can be automatically enlarged to the size of the camera by image processing. Therefore, it is not necessary to provide a complicated structure such as a movable platform and its control circuit in association with the camera unit 12, and the size and cost can be reduced. And the operability can be improved.

In addition, the image input device 10 according to the present invention is
When applied to the camera of the V-conference device, when transmitting image data from the transmission circuit to the display device via the communication line and the reception circuit, the data for all the images specified in the transmission capacity of the communication circuit etc. Since specific image data is output, the transmission capacity can be used effectively.
That is, when the transmitting side transmits all the image data and the receiving side enlarges the image, the transmission capacity increases in order to satisfy the image quality required for the enlarged image. Thus, by transmitting the data of the image enlarged to the size of the entire image in this manner, it is not necessary to set the transmission capacity unnecessarily large.

When the number of pixels in each area A is small, that is, when the original image is divided into a large number of divided images, unnecessary portions such as the background other than the specific subject can be reduced.

[0052]

Since the image input device of the present invention is configured as described above, the invention according to claim 1 is a CC as an image input means even when an object other than a subject to be watched is imaged.
An image including a specific subject can be instantly recognized without performing a pan / tilt operation of a D camera or the like, that is, a complicated operation such as vertical movement of a shooting range.

Further, according to the second aspect of the invention, even when a subject other than the subject to be watched is imaged, a pan / tilt operation of a CCD camera or the like as the image input means, that is, a complicated operation such as vertical movement of the photographing range. It is possible to easily recognize an image including a specific subject by selecting a limited image without performing.

According to the third aspect of the invention, even when an object other than the subject to be watched is imaged, a pan / tilt operation of a CCD camera or the like as the image input means, that is, a complicated operation such as vertical movement of the imaging range. For images containing the subject you want to gaze at, with a limited selection of images,
Camera control of focus, white balance, iris, etc. becomes possible and a clear image can be provided.

Further, according to the invention described in claim 4, even when an object other than the object to be watched is imaged, a pan / tilt operation of a CCD camera or the like as the image input means, that is, a vertical movement of the imaging range is complicated. You can enlarge a clear image with focus, white balance, iris, and other camera controls to an image containing the subject you want to gaze by selecting a limited number of images without any operation. Can be easily recognized.

Further, according to the invention of claim 5, even if a subject other than the subject to be watched is imaged, a pan / tilt operation of a CCD camera or the like as an image input means, that is, a complicated operation such as up and down movement of a photographing range. For images containing the subject you want to gaze at, with a limited selection of images,
Since a clear image in which camera control such as focus, white balance, and iris is performed can be stored as a still image, details that cannot be discriminated in real-time display can be precisely and easily recognized.

[Brief description of drawings]

FIG. 1 is a block diagram showing an electrical configuration of a first embodiment of an image input device of the invention.

FIG. 2 is an explanatory diagram for explaining an example of image division in the image input device of FIG.

FIG. 3 is an explanatory diagram for explaining an example of division of a captured image by the image division example of FIG.

FIG. 4 is a flowchart for explaining the operation of an image specifying unit of the image input device of FIG.

5 is an explanatory diagram for explaining a recognition operation of an image specifying unit of the image input device of FIG. 1. FIG.

6A and 6B are explanatory diagrams for explaining a selection operation of an image selection unit of the image input device of FIG.

FIG. 7 is a block diagram showing an electrical configuration of a second embodiment of the image input apparatus of the invention.

FIG. 8 is an explanatory diagram for explaining a function of the image input device in FIG.

[Explanation of symbols]

10 Image input device 11 subject 12 Camera section (image input means) 13 Image dividing unit (image dividing means) 14 Image designation section (image designation means) 15 Image specifying unit (image specifying means) 16 Image selection section (Image selection section) 17 Camera control unit (camera control means) 20 Image input device 28 Image Enlarging Section (Image Enlarging Means) 29 image storage unit (image storage means) Area A (divided image) P1 face P2 person P3 documents

Claims (5)

(57) [Claims] 1. An image input means for picking up an image of a subject and outputting the image, an image dividing means for dividing the image from the image input means into predetermined divided images, and a divided image divided by the image dividing means. From among these, an image designating unit that designates a single or a plurality of split images that include a part of a specific subject, and a designated split that is designated by the image designating unit from among the split images split by the image splitting unit. An image input device including: an image specifying unit that recognizes a single or a plurality of divided images that include an image and that includes an entire specific subject as specific divided images. 2. An image input means for picking up an image of a subject and outputting the image, an image dividing means for dividing the image from the image input means into predetermined divided images, and a divided image divided by the image dividing means. From among these, an image designating unit that designates a single or a plurality of split images that include a part of a specific subject, and a designated split that is designated by the image designating unit from among the split images split by the image splitting unit. Image specifying means for recognizing a single or a plurality of divided images including an image and the whole of a specific subject as a specific divided image, and an image selecting means for selecting the specific divided image recognized by the image specifying means. An image input device comprising: 3. An image input means for picking up an image of a subject and outputting the image, an image dividing means for dividing the image from the image input means into predetermined divided images, and a divided image divided by the image dividing means. From among these, an image designating unit that designates a single or a plurality of split images that include a part of a specific subject, and a designated split that is designated by the image designating unit from among the split images split by the image splitting unit. Image specifying means for recognizing a single or a plurality of divided images including an image and the whole of a specific subject as a specific divided image, and an image selecting means for selecting the specific divided image recognized by the image specifying means. And the focus, iris, white balance, etc. of the image input means are controlled based on the image data in the selected specific divided image selected by the image selection means. An image input device which comprises a control unit. 4. An image input means for picking up an image of a subject and outputting the image, an image dividing means for dividing the image from the image input means into predetermined divided images, and a divided image divided by the image dividing means. From among these, an image designating unit that designates a single or a plurality of split images that include a part of a specific subject, and a designated split that is designated by the image designating unit from among the split images split by the image splitting unit. Image specifying means for recognizing a single or a plurality of divided images including an image and the whole of a specific subject as a specific divided image, and an image selecting means for selecting the specific divided image recognized by the image specifying means. And the focus, iris, white balance, etc. of the image input means are controlled based on the image data in the selected specific divided image selected by the image selection means. Control means and an image input apparatus which comprises an image processing unit for enlarging the selected specific divided image selected by the image selection means. 5. An image input unit for capturing an image of a subject and outputting the image, an image dividing unit for dividing the image from the image input unit into predetermined divided images, and a divided image divided by the image dividing unit. From among these, an image designating unit that designates a single or a plurality of split images that include a part of a specific subject, and a designated split that is designated by the image designating unit from among the split images split by the image splitting unit. Image specifying means for recognizing a single or a plurality of divided images including an image and the whole of a specific subject as a specific divided image, and an image selecting means for selecting the specific divided image recognized by the image specifying means. And the focus, iris, white balance, etc. of the image input means are controlled based on the image data in the selected specific divided image selected by the image selection means. Control means and an image input apparatus characterized by comprising an image storing means for storing the selected specific divided image selected by the image selection means.