JP4798296B1 - Form - Google Patents

Abstract

Even when a form image is input by an image input device such as a camera that has an unstable form shooting state, the description position in the input image of each item is specified at high speed and processing is performed for each item. To provide a form processing apparatus and a recording medium that can be used.
An area symbol detection unit inputs a form that is surrounded by a start symbol and an end symbol that represent the direction and height of the processing target area surrounded by a certain undescribed area. The start symbol and end symbol are detected from the image, the region extraction unit extracts the processing target region using the position, height and direction of the start symbol and end symbol, and the processing unit describes the contents described in the processing target region. Is processed according to the type of the start symbol and end symbol, the start symbol and end symbol of the item are extracted from the input image of the form, the description position in the input image of each item is specified, and the processing for each item is performed. .
[Selection] Figure 1

Description

  The present invention relates to a form that indicates the type of information item and the entry position of the contents, and a form processing apparatus and form processing method for processing images after inputting the contents.

  Conventional form processing devices such as OCR, OMR, and various form readers that read characters and symbols written on recording media, detect photos, etc., use timing marks and ruled line information pre-written on forms. Thus, the inclination and scale of the entire form image are detected, and the description position in the input image of each item is specified using layout information registered in advance. FIG. 10A shows an example of a conventional form format.

Japanese Patent Laid-Open No. 5-159099

  However, since the conventional form processing apparatus assumes that the deformation of the input image is within a certain range, such as scanner input, when a form is photographed and input by a camera or the like, as shown in FIG. In addition, the image may be distorted due to a significant rotation of the input image or a shooting depression angle, and the extraction of timing marks and ruled lines may fail to extract item areas.

  Further, when entering the content, if the input character touches the ruled line, it is difficult to distinguish the ruled line from the entered content in the input image, and the ruled line extraction may fail.

  Further, when the entry position in each item is shown, it is difficult to distinguish between the notation indicating the entry position and the entry content in the input image, and the entry content processing may fail.

  The present invention solves the above-mentioned conventional problems, and even when a form image is input by an image input device such as a camera whose image capturing state is unstable, the position of each item in the input image is accurately described at high speed. It is an object of the present invention to provide a form processing apparatus and a recording medium that can identify each item and perform processing for each item.

Form of the present invention is constituted by entry sheet and the copy sheet, a form in which a copy sheet is read by the computer, the entry sheet and the copy sheet, the height and direction of the reading area when read in calculation machine has a terminator indicating the end position of the reading area when read start symbol and computer indicating a start position, the start symbol is indicates the height direction and the reading region of the reading area with respect to the start symbol, The copy sheet start symbol or end symbol is smaller than the entry sheet start symbol or end symbol .

  According to the form processing apparatus and the recording medium of the present invention, it is possible to specify an item area simply by extracting a start symbol and an end symbol of an item, and it is possible to cope with fluctuations in an input image at high speed and with high accuracy.

Configuration diagram of a form processing apparatus according to Embodiment 1 of the present invention Control schematic diagram of form processing apparatus in Embodiment 1 of the present invention Configuration diagram of a form according to the first embodiment of the present invention Explanatory drawing of the input form image in Embodiment 1 of this invention Explanatory drawing of the area | region symbol detection part in Embodiment 1 of this invention Explanatory drawing of the area | region extraction part in Embodiment 1 of this invention Explanatory drawing of the area | region extraction part in Embodiment 1 of this invention Explanatory drawing which shows the slip in Embodiment 1 of this invention. Explanatory drawing which shows the slip in Embodiment 1 of this invention. Explanatory drawing of conventional form processing apparatus and recording medium

  Embodiments of the present invention will be described below with reference to the drawings.

(Embodiment 1)
FIG. 1 is a configuration diagram of a form processing apparatus according to Embodiment 1 of the present invention.

In FIG. 1, an image input unit 101 inputs an image including a processing target area.
An area symbol detection unit 102 detects a start symbol, an end symbol, and a ruled line as necessary from the input image input by the image input unit 101.

  Reference numeral 103 denotes a region extraction unit that extracts a processing target region using the position, height, direction, and type of the start symbol and end symbol detected by the region symbol detection unit 102. A processing unit 104 processes the contents described in the processing target area extracted by the area extracting unit 103 according to the type of start symbol and end symbol. Reference numeral 105 denotes an output unit that outputs a result processed by the processing unit 104.

  Note that the functional blocks such as the image input unit, the region symbol detection unit, the region extraction unit, and the processing unit may be typically realized as an LSI that is an integrated circuit. The method of circuit integration is not limited to LSI. These may be individually made into one chip, or may be made into one chip so as to include a part or all of them.

FIG. 2 is a control schematic diagram of the form processing apparatus according to Embodiment 1 of the present invention. In an image input step S101, an image including a processing target area is input, and the process proceeds to S102. In area symbol detection step S102, a start symbol, an end symbol, and a ruled line are detected from the input image, and the position, height, direction, and type of the start symbol and end symbol are detected, and the process proceeds to S103.

  In region symbol verification step S103, the shape, combination, and positional relationship of the start symbol, the end symbol, and the ruled line are verified. If not, the process proceeds to S101, and if possible, the process proceeds to S104. In the region image extraction step S104, the processing target region image is extracted. If the processing target region image has been deformed, the process proceeds to S105, and if not, the process proceeds to S106, S107, S108 depending on the item.

  In the region image correction step S105, the processing target region image is corrected, and the process proceeds to S106, S107, and S108 depending on the item.

  In the character recognition step S106, a character recognition process is performed on the processing target area image, and the process proceeds to S109. In the symbol recognition step S107, a symbol recognition process is performed on the processing target region image, and the process proceeds to S109.

In the image storage step S108, the processing target area image is converted into image data, and the process proceeds to S109. In the output step S109, the results processed in the character recognition step S106, the symbol recognition step S107, and the image storage step S108 are output.
Each step may be described by a program that can be executed by a computer.

FIG. 3 is a configuration diagram of the form according to the first embodiment of the present invention. It consists of an entry sheet (FIG. 3 (a)) and a copy sheet (FIG. 3 (b)) and satisfies the following conditions.
The processing target area is surrounded by a start symbol, an end symbol, and a ruled line.
-The start symbol and the end symbol are surrounded by a certain undescribed region (white or background color).
The start symbol and the end symbol express the existence direction and height of the processing target area.
-The shape of the start symbol and end symbol differs depending on the processing content in the item column.
-The entry sheet shows the entry position in the processing target area and is not described on the copy sheet.
-Provide a certain unrecorded area between the start symbol, end symbol, entry position, and ruled line.
-The copy sheet start and end symbols are smaller than the corresponding entry sheet start and end symbols.

  For packaged items with slips, the entry sheet is kept at the client, and the contents are read at the collection / delivery place with only the copy sheet attached. Hereinafter, the details of the operation will be described by taking the input image of the copy sheet C of the form A shown in FIG. 4 as an example. When the copy sheet C is photographed by the camera in the image input unit 101, the image is input with the inclination θ degree and the resolution Ddpi as shown in FIG.

  The region symbol detection unit 102 extracts a connected closed region of the input image, performs shape matching with the start symbol and end symbol registered in advance as shown in FIG. Detect the position, height, direction, and type of symbols and end symbols. The detection result is shown in FIG.

  The start symbols LS1, LS2, and LS3 and the end symbols LE1, LE2, and LE3 were detected. In the copy sheet, since the start symbol and the end symbol are written smaller than the entry sheet, it is easy to detect as a connected closed region.

  In the area extraction unit 103, as shown in FIG. 6A, the form layout of the copy sheet C is registered in advance.

In the form layout, for each item, the processing content, the shape of the start / end symbol, the size of the processing area,
The entry position in the processing area is registered. With respect to the start symbol extracted by the region symbol extraction unit 102, an end symbol that is paired with the direction of the start symbol is specified. If it cannot be specified, the process returns to the processing of the image input unit 102.

  It is verified that LS1 and LE1, LS2 and LE2, LS3 and LE3 are a pair, and if the pair of start symbol and end symbol is known, the position of the processing target area is determined from the position, height and direction of the start symbol and end symbol. And cut out. FIG. 6B shows the result of specifying the position of the processing target area L1 of the start / end symbol pair LS1 and LE1.

Next, an angle correction method when the copy sheet C is read obliquely will be described. The inclination θ and the resolution D of the processing target area image can be obtained by the expressions shown in FIGS. 8 cm in FIG. 7B is the horizontal width of the postal area size stored in the form layout in FIG. Even if the start / end symbol of the pair can be detected, if the resolution is too low or too high, it is impossible to return to the processing of the image input unit 102. When the slope θ has a value, the processing target area image is rotationally corrected. FIG. 7C shows the processing target area image extracted.

The processing unit 104 performs processing corresponding to the item indicated by the start symbol for each extracted processing target region image. Since the processing target region image L1 shown in FIG. 7C is a zip code item from the form layout shown in FIG. 6A, the character image is used by using the information on the character description position in the registered item. To recognize characters.

Further, in the case of a selected symbol item, the selected symbol is identified using information on the symbol description position in the registered item. The output unit 105 outputs the processing results of these processing units 104.

  According to such a configuration, the start symbol and the end symbol are described in the processing target area of the form, and the start symbol and the end symbol of the item are extracted from the input image of the form by providing the area processing unit in the form processing apparatus. Since the position of each item in the input image is specified and processing is performed for each item, even when a form image is input with an image input device such as a camera where the form photographing state is unstable, each item is accurately processed at high speed. Processing for each item can be performed.

  Further, the effect of changing the size of the start symbol or the like between the entry sheet and the copy sheet will be described with reference to FIG. FIG. 8A shows an entry sheet on which an entry frame (solid line) is printed together with a start symbol and the like.

  FIG. 8B shows a copy sheet. Since the copy sheet (b) is directly below the entry sheet, the entry frame (solid line) of the entry sheet is at the position of the dotted line on the copy sheet. Actually, the dotted line is not printed.

  When “a” is entered so as to cover the character entry frame (solid line) of the entry sheet (a), “a” is copied to the position of FIG. 8B on the copy sheet.

  At this time, in the copy sheet, the start symbol is smaller than the start symbol of the entry sheet and is printed away from the entry frame, so that the entered character does not cover the start symbol. For this reason, character recognition can be performed easily.

  Furthermore, the case where the character in the start symbol of an entry sheet and a copy sheet is printed is described with reference to FIG. In FIG. 9, as shown in the entry sheet (a), the start symbol has a postal mark, but is not printed on the copy sheet (b).

  For this reason, when printing on an entry sheet, the user can grasp the contents of entry, and when reading a copy sheet with a computer, since there is no symbol, it is easy to detect the start symbol by image processing. .

  In the present embodiment, the copy sheet is the processing target, but the entry sheet may be the target. When the entry sheet is a processing target and the entry sheet is input as an image, a process of deleting the image at the entry position indicated by using the entry position information in the item may be added.

The start / end symbols, ruled lines, and notation indicating the entry position may be printed in different colors.
The shape of the start / end symbol is not limited to this. A plurality of start / end symbols may be arranged at the outer peripheral position of the processing target region in order to cope with a case where the processing target region has a complicated shape.

The form, form processing apparatus, and form processing method according to the present invention are:
It has a high-speed and high-performance image processing function and is useful as an image processing method.
In addition, it has a rapid information data conversion function, and is useful as a method for creating an information database.

DESCRIPTION OF SYMBOLS 101 Image input part 102 Area | region symbol detection part 103 Area | region extraction part 104 Processing part 105 Output part

Claims (2)

The form is composed of an entry sheet and a copy sheet, and the copy sheet is read by a computer. The entry sheet and the copy sheet are:
Has a terminator indicating the end position of the read-area when read by a computer start symbol indicating the height and direction and the start position of the read-area when read in and computer,
The start symbol indicates the direction and height of the read-region of the read-region for the start symbol, that start symbol or terminator of the copy sheet is less than the start symbol or terminator of the entry sheet A feature form. The form according to claim 1, wherein the start symbol and the end symbol have different shapes.

Images