JP2003178071A - Document management system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a document management system capable of efficiently classifying document images by extracting an areal image of a drawing or a photograph, and using the same as an attribute. <P>SOLUTION: When an image input part 1 inputs the document image, an area discriminating processing of the document image is executed by an area discriminating means 2. An image cutout part 3 classifies a drawing area or a photograph area among the areas divided into a sentence area, a table area, a ruled line, the drawing area and the photograph area by the area discriminating part 2, as the 'drawing area', and a partial image of the area is cut out. A document image management part 4 adds the cutout partial image to the management information of the document image as the attribute information, and stored in a document image storing part 5. When two or more drawing areas or the photograph areas exist in the inputted document image, the image cutout part 3 can cut out all of the drawing areas as the partial image, or representative drawing area or the photograph area among two or more drawing areas or the photograph areas, as the partial image. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document image management system for adding and managing one or more attributes to an input document image.

[0002]

2. Description of the Related Art Generally, in a document image management system such as an electronic document filing apparatus, a method of giving one or more attributes to an input document and managing it is often adopted. Conventionally, the addition of attributes to document images is
Information that the document management system can easily acquire, such as the date and time when the document image was registered and the registrant who registered the image document, is used as additional information. In addition to the information that can be easily acquired by the document management system, the operator directly inputs a character string or selects an attribute to be added from a pre-registered attribute list as additional information of the attribute to the document image. In general, it is often done manually. There is also a method of sorting documents by type before inputting and inputting as a document image collectively for each type. As a method of inputting and managing such a document image, Japanese Laid-Open Patent Publication No. 10-240958 discloses that a fixed form is identified, information included in the document is read, and the read information and the document are linked. A management information extracting apparatus and method for automatically extracting information from a document by management and extracting management information from an image that can reduce the work of an operator are described.

[0003]

However, a general manual method such as an operator inputting a character string or selecting an attribute to be added, a method of sorting documents by type before inputting, and then collectively documenting The method of inputting as an image is inefficient because the load on the operator becomes large as the number and type of document images increase. Further, in the invention described in Japanese Patent Laid-Open No. 10-240958, since the attribute to be added to the document is obtained by reading the information described on the document, the actually required attribute is described on the corresponding document. It is not always the case that accurate attributes cannot be extracted. In such a case, it is necessary to manually add a desired attribute after inputting the document. In order to reduce the work of the operator at that time, the attributes given to each fixed form of the input document are defined, and the attributes given to the document image are determined by identifying which form the input document image is. Although a method of deciding is also proposed, the document image input to the document image management system is not always a fixed form.

Furthermore, by reducing the load of the operator's manual work and giving an attribute using the area information of the input document image, an appropriate attribute can be given to the input document image even if it is not a fixed form. Document management systems capable of automatically classifying have also been proposed. In this document management system, since the attribute is given by using the area information of the input document image, the information can be input without writing the information at a specific position on the document image or using a fixed form. Appropriate attributes are given to the created document images to automatically classify them. The information used as an attribute in such a case is based on area identification results, layout information, character recognition results, barcode recognition result information, etc. In addition, logos, graphs, other figures and photographs contained in the document image may occupy an important position in classifying the document image. For example, if documents having similar figures or photographs are used for classification, it is possible to efficiently classify them without examining the contents in detail.

Therefore, a first object of the present invention is to extract an area image of a figure or photograph when the figure or photograph exists in the inputted document image and use it as an attribute to input the document. An object of the present invention is to provide a document management system capable of efficiently classifying images. A second object of the present invention is to add an attribute such as an extracted figure or photograph to a document image input by an operator later, and to automatically classify the document image based on the added attribute. An object of the present invention is to provide a document management system that can reduce the load of manual work.

[0006]

According to a first aspect of the present invention, a document image acquiring unit for acquiring a document image and a document image acquired by the document image acquiring unit have one or more drawing areas or photograph areas. Image area recognition means for recognizing whether or not to do so, and when the image area recognition means recognizes that one or more drawing areas or photograph areas exist in the document image, the recognized drawing area or photograph area is used as drawing area information. Area information extracting means for extracting as the attribute, attribute adding means for adding the drawing area information extracted by the area information extracting means to the document image as an attribute, and a document image storing the document image to which the attribute adding means adds the attribute By including the storing means, the first and second objects are achieved.

According to a second aspect of the invention, in the first aspect of the invention, the area information extracting means extracts all of the plurality of drawing areas or photograph areas recognized by the image area recognizing means as drawing area information. Thus, the first and second objects are achieved. According to the invention of claim 3,
In the invention according to claim 2, further comprising area information selecting means for prompting selection of drawing area information to be added as the attribute when the area information extracting means extracts a plurality of drawing area information, and the attribute adding means, By adding the drawing area information selected by the area information selecting means to the document image as an attribute, the first and second objects are achieved.

According to a fourth aspect of the present invention, in the invention according to any one of the first, second and third aspects, the feature of the single or plural figure area information extracted by the area information extracting means. Further comprising a feature extracting unit for extracting the feature, and the attribute adding unit adds the feature extracted by the feature extracting unit to the document image as an attribute,
The first and second objects are achieved.

According to a fifth aspect of the invention, in the fourth aspect of the invention, the document image storage means further comprises a document image classification means for classifying the document images based on the features extracted by the feature extraction means. Achieves the first and second objects by storing the document image for each classification by the document image classification means.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, preferred embodiments of the present invention will be described in detail with reference to FIGS. FIG. 1 is a block diagram showing a schematic configuration of the document management system according to the first embodiment. As shown in FIG. 1, the document management system includes an image input unit 1 for inputting a document image using an image input device such as a scanner, and the input document image in one or more areas such as sentences, diagrams, photographs and tables. The area identification unit 2 that divides and identifies the image or the area of the photograph based on the area identification result by the area identification unit 2, and the image or image area cut out by the image cutout unit 3 is a document. A document image management unit 4 for adding an attribute to an image and a document image storage unit 5 for storing an input document image are provided.

FIG. 2 is a flow chart showing a processing procedure of an image clipping process and an attribute adding process of an input document image by the document management system of the first embodiment. First, when the image input unit 1 inputs a document image (step 2
01), the area identifying unit 2 executes the area identifying process of the input document image (step 202). Here, the area identification processing by the area identification unit 2 will be described. For example, the input document image is compressed into a binary compressed image, a rectangle circumscribing the black pixel connected component of the binary compressed image is extracted, and the rectangle is divided into a character rectangle and other rectangles. There is a method of extracting character areas and other areas by classifying the rectangles and integrating them. The rectangles extracted here include one or more character strings, ruled lines, tables, diagrams, graphs, photographs, and other rectangles having various attributes. In this way, the document image is divided into a text area, a table area, ruled lines, a drawing area, a photograph area, and the like. By adding information about the type and number of these areas to the document image as an attribute, it becomes possible to search later in a document management system using the number of figures and tables included in the document image. ing.

Next, the image cutout unit 3 includes the area identification unit 2
Of the areas divided into text areas, table areas, ruled lines, figure areas, photo areas, etc., figure areas or photo areas are grouped together as "figure areas", and areas classified as "figure areas" Cut out a partial image of (Step 20
3). Then, the document image management unit 4 adds the clipped partial image to the management information of the document image as the attribute information of the document image (step 204), and stores it in the document image storage unit 5 together with the attribute information about the document image ( Step two
05). Thus, in the document management system, when the user refers to the list of the management information of the managed document images, the user can easily obtain a desired image area by simply glancing at the drawing area or the photograph area cut out from the document image displayed together. The document image of can be selected and acquired.

If there are a plurality of drawing areas or photograph areas in the input document image, the image cutting unit 3 may cut out all the drawing areas as partial images. In this case, the document image management unit 4 adds all the cut out partial images to the management information of the document image as attribute information regarding the document image and stores it. In this way, the document image management unit 4 adds the attribute information relating to all the drawing areas or the photo areas described in the document image cut out by the image cutout unit 3 to the management information.
Even if the input document image contains the same drawing area or photograph area and it is not possible to clearly distinguish it with only one kind of partial image of the drawing area or photograph area, the user can easily request it. The document image of can be selected and acquired.

Further, the image cut-out section 3 determines whether the input document image has a plurality of drawing areas or photograph areas.
Of these plural drawing areas or photograph areas, a representative drawing area or photograph area may be cut out as a partial image. In this case, the document image management unit 4 adds the partial image cut out as a representative one to the management information of the document image as attribute information about the document image and stores it. If all drawing areas or photo areas are added as attribute information, the amount of memory or disk consumption may increase when the number of attribute information is large. Therefore, in order to prevent an increase in the amount of information to be stored, one or a plurality of typical drawing areas or photo areas are selected and only the selected ones are added as attribute information, thereby consuming memory and disk. The amount can be reduced. As a typical method of selecting a drawing area or a photograph area, there is a method of displaying a list of drawing areas or photograph areas and giving an instruction for selection by an operator or a user based on the displayed list.

As described above, in the document management system of the first embodiment, the image portion of the drawing area or the photograph area is added as an attribute value to the input document image, so that it is easy to understand. It is possible to characterize a typical document image. Further, in the document management system of the first embodiment, the image portions of all the drawing areas or the photograph areas are added as attribute values to the input document image, so that the same drawing area or photograph area is added. It is possible to easily specify the document image having the image portion of, and facilitate mutual identification between the document images. Further, in the document management system of the first embodiment, an image portion of a predetermined drawing area or photograph area is selected for an input document image, and the selected image portion is added as an attribute value. Therefore, it is possible to reduce the amount of memory and disk consumption.

Next, a second embodiment will be described.
FIG. 3 is a block diagram showing a schematic configuration of the document management system according to the second embodiment. As shown in FIG.
The document management system according to the second embodiment further includes an image feature extraction unit 6 that extracts the features of the drawing region or the photo region cut out by the image cutout unit 3 in the document management system according to the first embodiment. Has become. Below, the first
The same parts as those in the embodiment will be described using the same numbers. The document image management unit 4 further adds the image features extracted by the image feature extraction unit 6 as attribute information of the document image, classifies the document images by type, and stores the document images in the document image storage unit 5.

FIG. 4 is a flow chart showing the processing procedure of the image feature extraction processing and the attribute addition processing of the input document image by the document management system of the second embodiment. First, when the image input unit 1 inputs a document image (step 3
01), the area identifying unit 2 executes the area identifying process of the input document image (step 302). The area identification processing by the area identification unit 2 is the same as the processing described in FIG. 2, and thus the description thereof is omitted here. The image cutout unit 3 classifies the figure region or the photo region out of the respective regions divided into the text region, the table region, the ruled line, the figure region, the photo region, etc. by the region identification unit 2 and classifies them as a “figure region”. Then, the partial image of the area classified as the "drawing area" is cut out (step 303).

The image feature extraction unit 6 cuts out the drawing region or the photographic region by the image cutout unit 3, and then executes the feature extraction processing on the cut out partial image (step 304). For example, as the feature extraction processing, based on the color of each pixel forming the prepared image, it is specified which area in the color space each pixel belongs to in advance, and each area is identified. There is an image feature extraction method of generating a color histogram of an image based on the number of pixels belonging to and extracting the generated color histogram as a feature amount representing a feature of the image. In addition, when extracting the histogram of the color of the image generated based on the number of pixels belonging to each area as the feature amount representing the feature of the image, the number of pixels belonging to each area is determined according to the degree of color accumulation between areas. And an image feature extraction method that extracts the histogram of the generated color as a feature amount representing the feature of the image based on the number of pixels belonging to each region after the accumulation.

Then, the document image management unit 4 adds the image features extracted by the image feature extraction unit 6 to the management information of the document image as attribute information of the document image (step 30).
5), it is stored in the document image storage unit 5 together with the attribute information about the document image (step 306). Thus, in the document management system, when the user refers to the list of the management information of the managed document images, the user can easily glance at the image features extracted from the document images displayed together with the desired document image. You can select and get. Also, since the image features are extracted and added as attribute information,
Compared to the case where the image itself is added as the attribute information, the consumption of memory and disk can be significantly reduced.

FIG. 5 is a flow chart showing the processing procedure of the document image classification processing and the attribute addition processing of the input document image by the document management system of the second embodiment. First, when the image input unit 1 inputs a document image (step 5)
01), the area identifying unit 2 executes the area identifying process of the input document image (step 502). The area identification processing by the area identification unit 2 is the same as the processing described with reference to FIG. The image cutout unit 3 classifies the figure region or the photo region out of the respective regions divided into the text region, the table region, the ruled line, the figure region, the photo region, etc. by the region identification unit 2 and classifies them as a “figure region”. The partial image of the area classified as the "drawing area" is cut out (step 503). The image feature extraction unit 6 cuts out the drawing region or the photographic region by the image cutout unit 3, and then executes the feature extraction processing on the cut out partial image (step 504). The image feature extraction processing by the image feature extraction unit 6 is the same as the processing described in FIG. 4, and thus the description thereof is omitted here.

The document image management unit 4 adds the image features extracted by the image feature extraction unit 6 to the document image management information as attribute information of the document image (step 505). Here, the document image management unit 4 adds the attribute information about the document image and classifies the document images to be stored by a statistical method using the image feature as the attribute information (step 506). The classification of document images to be saved is
Since the processing is performed based on the image feature, it is possible to automatically perform the group classification according to predetermined requirements such as grouping the document images having similar image features.
Then, the grouped document images are stored in the document image storage unit 5 for each classification (step 507). Thus, when the user searches for a document image, the search range can be limited by referring to the document image for each classification grouped in the document image storage unit 5, and an efficient document management system can be obtained. Can be realized.

As described above, in the document management system of the second embodiment, a partial image of a drawing area or a photographic area is cut out from an input document image, and the feature of this partial image is extracted and input as an attribute value. Since it is added to the created document image and saved / managed, the consumption of memory and disk can be significantly reduced compared to the case where the partial image itself is added as an attribute value. Further, in the document management system of the second exemplary embodiment, a partial image of a drawing area or a photographic area is cut out from an input document image, and predetermined classification is executed using image features extracted as features from this partial image, Since the document images are stored and managed for each classification, if the barcode is added to a specific area as the area information in the document image, the barcode information can be used to easily document the document. Images can be classified, and efficient document image storage / management can be realized.

[0023]

According to the first aspect of the present invention, when the image area recognizing means recognizes that a document image has one or more drawing areas or photograph areas, the recognized drawing area or photograph area is used as drawing area information. Since the area information extracting means for extracting and the attribute adding means for adding the drawing area information extracted by the area information extracting means to the document image as attributes are provided, retrieval, management, etc. can be performed by efficient characterization of the document image. Can be done easily.

According to the second aspect of the present invention, since the area information extracting means extracts all of the plurality of drawing areas or photograph areas recognized by the image area recognizing means as drawing area information, images of similar drawing areas or photograph areas are extracted. It is possible to facilitate the discrimination between the document images having the.

According to the third aspect of the invention, when the area information extracting means extracts a plurality of drawing area information, it further comprises area information selecting means for urging selection of the drawing area information to be added as an attribute. It is possible to select an image in an arbitrary drawing area or photo area and add it as an attribute value to the document image, and it is possible to reduce the capacity consumption of the memory or the disk.

According to the fourth aspect of the present invention, since the feature information extracting means for extracting the feature of the single or plural drawing area information extracted by the area information extracting means is further provided, the drawing area is added to the input document image. Alternatively, the characteristics of the photo area can be extracted and added as an attribute value, and the capacity consumption of the memory or disk can be significantly reduced.

According to the fifth aspect of the present invention, since the document image classification means for classifying the document images based on the features extracted by the feature extraction means is further provided, the bar code information or the like is assigned to the specific area as the drawing area information. In the case as described, classification can be performed based on this barcode information, and efficient document image management can be realized.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a document management system according to a first embodiment.

FIG. 2 is a flowchart showing a processing procedure of an image clipping process and an attribute adding process of an input document image by the document management system of the first embodiment.

FIG. 3 is a block diagram showing a schematic configuration of a document management system according to a second exemplary embodiment.

FIG. 4 is a flowchart showing a processing procedure of image feature extraction processing and attribute addition processing of an input document image by the document management system of the second embodiment.

FIG. 5 is a flowchart showing a processing procedure of a document image classification process and an attribute addition process of an input document image by the document management system of the second embodiment.

[Explanation of symbols]

1 Image input section 2 Area identification part 3 Image cutout section 4 Document image management section 5 Document image storage 6 Image feature extraction unit

Claims (5)

[Claims] 1. A document image acquisition unit for acquiring a document image, and an image region recognition unit for recognizing whether or not a document image acquired by the document image acquisition unit has one or more drawing regions or photograph regions. Area information extracting means for extracting the recognized drawing area or photo area as drawing area information when the image area recognizing means recognizes that one or more drawing areas or photograph areas exist in the document image; An attribute adding unit that adds the drawing area information extracted by the extracting unit to the document image as an attribute, and a document image storing unit that stores the document image to which the attribute adding unit has added the attribute are provided. Document management system. 2. The document management system according to claim 1, wherein the area information extracting means extracts all of the plurality of drawing areas or photograph areas recognized by the image area recognizing means as drawing area information. 3. When the area information extracting means extracts a plurality of drawing area information, the area information selecting means further includes area information selecting means for urging selection of the drawing area information to be added as the attribute, and the attribute adding means includes the area information. 3. The document management system according to claim 2, wherein the drawing area information selected by the selection means is added to the document image as an attribute. 4. The apparatus further comprises a feature extraction means for extracting a feature of the single or plural drawing area information extracted by the area information extraction means, wherein the attribute addition means also uses the feature extracted by the feature extraction means as an attribute. The document management system according to any one of claims 1, 2, and 3, which is added to a document image. 5. A document image classifying unit that classifies the document images based on the features extracted by the feature extracting unit, wherein the document image storing unit includes the document images for each classification performed by the document image classifying unit. The document management system according to claim 4, wherein the document management system stores the document.