JPH05128302A - Method and device for character recognition - Google Patents

Abstract

PURPOSE:To properly display character recognition results in accordance with the character recognition volume. CONSTITUTION:The picture document read in from a scanner is binarized, and an obtained image is displayed in an image window 11. When an area to be subjected to character recognition is designated on the displayed image, a character picture included in this area is segmented character by character to calculate the number of characters and the number of rows in the area. The size of a result window 13 where recognition results should be displayed is obtained based on these calculated numbers of characters and rows, and this window is opened. Finally, character recognition for the character picture in the designated area is executed, and results are displayed in the result window 13.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition method and an apparatus therefor, and more particularly to a character recognition method and an apparatus therefor for optically reading and recognizing a character printed or handwritten on paper.

[0002]

2. Description of the Related Art A conventional optical character recognition device (hereinafter referred to as OC
The character recognition process in (R: Optical Character Reader) is generally executed by the procedure shown in the flowchart of FIG.

That is, after the image original is read from the scanner in step S501 and an analog electric signal output is obtained, this is binarized in step S502 to be converted into binary digital data (hereinafter referred to as an image). Next, in step S503, an image window screen for displaying an image is opened on the display provided in the device, and the image is displayed, for example, as shown in FIG. Normally, a plurality of window screens can be simultaneously displayed on the display, but FIG. 16 shows an example in which the image window 101 is displayed in the entire display screen.

In step S504, while looking at the image displayed on the display, the device operator selects an area to be recognized (hereinafter referred to as a recognition target area), for example, the area indicated by 102 in FIG. Specify with a pointing device such as a mouse. Step S50
In 5, the result window for displaying the recognition result of the character image included in the recognition target area is opened. Further, in step S506, a character cutout process for extracting the character images in the designated recognition target area one by one is executed, and in step S507, the recognition process is executed for the cut out image of each character. Finally step S5
At 08, the recognition result is displayed in the result window.

In addition to the general character recognition processing procedure as described above, in the case where a photograph, a drawing, a headline, etc. are included in the input image original and the recognition target area cannot be cut out in a rectangle designated by the mouse. In addition, before the area designation by the mouse in step S504, the device operator performs an operation such as erasing an unnecessary portion from the image displayed in the image window 101.

For example, when it is desired to recognize the character image included in the area 103 in the image displayed in the image window 101 shown in FIG. "Static induction"
It is inconvenient because it includes the heading. For this reason, the device operator performs an operation of deleting the "electrostatic induction" portion from the displayed image to obtain an image like 103 'shown in FIG. 18, and then designates a rectangular area with the mouse. It was

[0007]

However, in the above-mentioned conventional example, since the result window is displayed based on the predetermined initial size, the number of characters or lines in the recognition target area does not correspond to the size of the result window, which is wasteful. There was a drawback that it was difficult to display the part and the recognition result was difficult to see.

In this regard, the result window 104 and the result window 1 of various sizes shown in FIGS.
A specific description will be given with reference to the recognition result displayed in 04. The recognition results shown in these figures are obtained from the character image included in the recognition target area 102 designated from the image displayed in the image window 101 shown in FIG. 19 to 21, 104 is a result window, and 105 is a mode setting window for setting various modes necessary for the recognition operation.

In the case shown in FIG. 19, the result window 10 with respect to the number of characters and the number of lines in the recognition target area 102 is displayed.
Since the size of 4 is too large, there is a wasteful part. On the other hand, in the case shown in FIG. 20, the result window 1
Since the size of 04 is too small, a forced line break due to the window size limitation is added to the original line break, which makes it very difficult to see. To compensate for the above drawbacks, a method of ignoring the original line breaks and inserting only line breaks matching the window size can be considered as shown in FIG. 21, but in this case,
Since the line feed position is different from the original sentence, it still feels uncomfortable visually, and the inconvenience that it is difficult to confirm and correct the recognition result remains.

Further, when it is desired to recognize an area other than a rectangle from an input image document including a photograph, a figure, a headline, etc., extra work is required to delete an unnecessary portion. The problem that it takes time to input,
There was always the possibility of an operational mistake in deleting the unnecessary part by mistake while deleting the unnecessary part. The present invention has been made in view of the above conventional example, and an object of the present invention is to provide a character recognition method and apparatus capable of displaying a recognition result in an appropriate size corresponding to the size of a character image included in a recognition target area. I am trying.

[0011]

In order to achieve the above object, the character recognition method of the present invention comprises the following steps. That is, it is a character recognition method for recognizing a character drawn on the image original by reading the image original, and an input step of reading the image original, and a first step of displaying the image original read by the input step on a screen. A first display step of displaying in a window, a designation step of designating an area where character recognition is desired from the image original displayed by the first display step, and a character included in the area designated by the designation step Is calculated for each character to calculate the number of characters included in the area, and the size of the second window screen for displaying the character recognition result is determined based on the number of characters calculated by the calculating step. A window opening step of opening the second window above, and a recognition step of recognizing the characters contained in the area one by one Comprising a character recognition method characterized by having a second display step of displaying the recognition result by said recognizing step in the second window.

According to another aspect of the invention, there is provided a character recognition device for reading an image original and recognizing characters drawn on the image original, wherein the input means reads the image original and the character read by the input means. First display means for displaying the image original in a first window on the screen, designating means for designating an area for character recognition from the image original displayed by the first display means, and by the designating means A calculating unit that calculates the number of characters included in the region by cutting out characters included in the specified region one by one, and a second window that displays the character recognition result based on the number of characters calculated by the calculating unit. A window opening unit that determines the size of the screen and opens the second window on the screen, and a sentence included in the area Each comprising a character recognizing unit, a character recognition apparatus characterized by a second display means for displaying the recognition result by the recognition means to the second window.

[0013]

With the above arrangement, the present invention designates an area in the input image original where character recognition is desired, and displays the character recognition result based on the number of characters included in the area.
Works to determine the size of the windows screen and open it.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described in detail below with reference to the accompanying drawings.

[General Description of Apparatus (FIG. 1)] Here, a character recognition apparatus commonly used in several embodiments described below will be described.

FIG. 1 is a block diagram showing the configuration of an optical character recognition device (OCR) which is a typical embodiment of the present invention. In FIG. 1, reference numeral 1 denotes a scanner that irradiates an image original with light and reads the reflected light to output an analog electric signal.
Reference numeral 2 is a binarization circuit that digitally converts the electric signal output from the scanner 1 into a binary signal corresponding to black and white, and 3 is a mouse that is a pointing device for inputting coordinates such as area designation on an image window. 4 is mouse 3
A mouse interface circuit for transmitting signals from other devices to other device components, 5 is a CPU for controlling the entire device and performing character cutout processing, recognition processing, etc., 6 is a CPU
A ROM storing a control program executed by 5, various processing programs, a recognition dictionary, and the like, and 7 is an RA used as a work area for developing a character image or character recognition processing.
M and 8 are displays for displaying input images and recognition results, 9 is a display interface circuit, and 10 is a CPU bus for connecting each device constituent element.

[First Embodiment (FIGS. 2 to 4)] Here, a character recognition process executed by using the character recognition device having the above configuration will be described with reference to a flow chart shown in FIG. Note that, here, the same processing steps as those already described in the conventional example are designated by the same step reference numerals, and description thereof will be omitted.

In steps S501 to S504, input of an image original, binarization processing, image display, and designation of a recognition target area by a mouse are performed according to a conventional technique.

Next, in step S10, character cutting processing is performed from the designated recognition target area, and a rectangular area corresponding to each character is taken out. Then, in S15, the maximum value of the number of characters and the number of lines in the recognition target area are calculated based on the result of the character cutting process. From these values, the size of the result window that is necessary and unnecessary for displaying the recognition result character is determined, and in step S20, the result window according to the calculated size is opened.

In steps S507 to S508, recognition processing is executed according to the conventional technique, and the result is displayed in the result window.

By the above processing, it is possible to obtain an appropriate display of the relationship between the size of the result window and the display character amount of the character recognition result as shown in FIG. In FIG. 3, 11 is an image window, 13 is a result window, and 14 is O.
It is a mode setting window for setting various modes necessary for the operation of the CR. Also here the result window 1
The character displayed in No. 3 is obtained from the character image included in the recognition target area 102 designated from the image displayed in the image window 101 shown in FIG.

In the above description, an example of the process for a horizontally written document is shown, but it goes without saying that the same process can be applied to a vertically written document.

Further, in the above description, the multi-window screen is an overlap type multi-window in which the windows are overlapped, but for example, a tile-type multi-window in which the windows are not overlapped as shown in FIG. 4 may be used. Of course it is possible. In FIG. 4, 15
Is a processing mode setting window, 16 is an image window, and 17 is a result window. Here, the height of the result window is fixed, but the width is determined according to the maximum number of characters in the recognition target area.

Further, it goes without saying that each window can be freely moved, resized (enlarged / reduced, etc.), placed in front and placed in the rear by operating the mouse.

Therefore, according to the present embodiment, the size of the result window is determined based on the maximum value of the number of characters and the number of lines included in the designated recognition target area.
A result window suitable for the display character amount of the character recognition result can be obtained, and an appropriate display can be performed without any discomfort.

[Second Embodiment (FIGS. 5 to 11)] In the first embodiment, the optimum size of the result window is determined in accordance with the maximum number of characters and the number of lines included in the designated recognition target area. The processing has been described. In the present embodiment, a process capable of designating a recognition target area having a more complicated shape than a simple rectangle will be described with reference to the flowchart shown in FIG. Here, also here, the same processing steps as those already described in the conventional example are designated by the same step reference numerals, and description thereof will be omitted.

In steps S501 to S503, input of an image original, binarization processing, and image display are performed according to the conventional technique.

Next, in step S30, the recognition target area is specified using the mouse 3. In the present embodiment, the mouse 3 is operated to draw a line segment on the image window screen, the line segments are connected to form a polygon, and the region surrounded by the polygon is regarded as the recognition target region. The operation of designating the recognition target area will be specifically described with reference to the display screen examples shown in FIGS.

FIG. 6 is a diagram showing a state in which the image original input from the scanner 1 is displayed in the image window 41 after the processing of step S503 is completed. Further in FIG.
The line segment 42 and the line segment 43 have already been drawn, indicating that the polygon for designating the recognition target area is being drawn. The line segment is specified by the mouse 3 and the cursor 4
After moving 0 to the desired position, press the mouse 3 button 1
By clicking twice, the start point or the end point of the line segment that is a component of the polygon is determined.

In FIG. 7, drawing of a polygon is advanced as compared with FIG.
It shows how to determine the last line segment. In FIG. 7, the mouse 3 is moved while further moving the cursor 40 to a desired position.
The button is continuously clicked once to draw line segment 44, line segment 45, and line segment 46. In order to complete the polygon, the cursor 40 is moved to a desired position and then the button of the mouse 3 is clicked twice (double click). As a result, for example, in the case shown in FIG. 7, the position of the cursor 40 and the starting point 51 at which the polygon is started are connected by a straight line, and the polygon 61 shown in FIG. 8 is completed. It is regarded as a recognition target area.

When the designation of the recognition target area is completed in this way, the process proceeds to step S35, where character cutting processing is executed in the designated area to extract a rectangular area for each character. Here, a more specific description will be given with reference to FIGS. 9 and 10. It should be noted that although many methods are already known as known techniques for character extraction, in the present embodiment, character extraction processing is performed using a histogram of black pixels obtained by projection. Needless to say, other cutting methods such as contour line tracking may be used.

In order to simplify the explanation, here, the designated recognition target area is a rectangular area 71 as shown in FIG.
And Here, considering an xy coordinate system having an x-axis in the horizontal direction and a y-axis in the vertical direction as shown in FIG. 9, the coordinates of this rectangular area are a ₀ <x <a ₁ and b ₀ <y <. It can be expressed as b ₁ .

First, when a black pixel is projected in the x direction within the rectangle 71, a histogram 72 is obtained in the y axis direction. According to this histogram 72, Y _S1 , Y _E1 , Y
The y coordinate values of the start point and the end point of each line are calculated as _S2 , Y _E2 , ....

Next, for each row, that is, for the nth row, in the rectangle given by a ₀ <x <a and Y _Sn <y <Y _En , as shown in FIG. Histogram 73 is obtained in the x-axis direction.
Then, according to this histogram 73, X _S1 , X _E1 ,
The x coordinate values of the start point and the end point of each character x _S2 , X _E2 , ... _Are calculated.

The above concept can be applied to the polygonal recognition target area shown in FIG. In the case of the polygon 74 of FIG. 11, the coordinates of the area surrounded by the polygon are X _S (y) <x <X _E (y) (1) and X _S (x) <y <Y _E (x ) ... (2) is represented.

Therefore, by projecting in the x direction and the y direction only within the range satisfying the conditions of the expressions (1) and (2), it is possible to cut out a line and a character in a desired area.

Finally, in steps S507 to S508, the recognition process is executed on the cut out character image according to the conventional technique, and the result is displayed in the result window.

Therefore, according to the present embodiment, since the polygon to be freely combined with a plurality of line segments becomes the recognition target area, the recognition target area having a complicated outline which cannot be simply specified with the mouse is specified. be able to. In addition, even if there are photographs, figures, captions, etc., it is possible to easily remove them and select the recognition target area.

Although the mouse is used as the pointing device in the above description, various other pointing devices can be used. Also, it has been explained that a single click of a mouse button is used to confirm a line segment when forming a polygon, and a double click is used to confirm a polygon, but the definition of these clicks can be changed. Alternatively, the line segment or the polygon may be fixed when the mouse button is released.

[Third Embodiment (FIGS. 12 to 14)] In the second embodiment, a case will be described in which a plurality of line segments are combined to form a polygon, and an area surrounded by the polygon is used as a recognition target area. However, in the present embodiment, a case will be described in which simple rectangles are combined to form a recognition target area. When compared with the second embodiment, the feature of the present embodiment lies in the procedure for forming the recognition target area, and the other processing is common, so here only the formation procedure will be described with reference to FIGS. 12 to 14. explain.

As an area to be recognized, consider the area 103 shown in FIG. 17 described in the conventional example. First, the rectangular area 81 shown in FIG. 12 is designated by the mouse 3. Rectangular area 81
First, move the mouse cursor to the upper left corner of the rectangular area 81 and click the mouse button once, then drag the mouse to the lower right corner of the rectangular area 81 and click the mouse button again. Determine. Next, the mouse 3 is similarly operated to operate the rectangular area 82.
Is specified. When a plurality of rectangular areas are successively designated by the mouse in this way, in the present embodiment, the designated rectangular areas are combined to form a polygon. For example, in FIG.
In the case of FIG. 1, the line segment 83 and the line segment 84 in the portion where the rectangular regions 81 and 82 overlap each other are ignored, and as a result, as shown in FIG.
A polygon 85 as shown in 3 is regarded as a recognition target area. Thus, the recognition target area 86 as shown in FIG. 14 is finally obtained. It is represented by.

In other words, it can be said that the synthesis of the rectangular area means leaving only the outermost portion of the line segment which constitutes the designated rectangle. Such a combining process is executed every time a rectangular area having an overlap is designated, and only the outermost line segment element is always left as the boundary line of the recognition target area.

Therefore, according to the present embodiment, a polygon formed by designating a plurality of rectangular areas with the mouse and synthesizing them can be the recognition target area.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of one device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0045]

As described above, according to the present invention, the size of the window for displaying the character recognition result is determined and opened based on the number of characters included in the area in which character recognition is desired. This has the effect of obtaining a window of an appropriate size for the result. Therefore, there is an advantage that the window is not too large and an unnecessary portion is not generated on the screen, and that the window is not too small and the character recognition result is easy to see, so that the recognition result can be easily confirmed and corrected.

Also, since the operator can specify the area in which character recognition is desired with a polygon of a desired shape, even an area having a shape that must be divided into several times and specified can be specified at once. The operation efficiency is improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a character recognition device that is a typical embodiment of the present invention.

FIG. 2 is a flowchart showing a character recognition process according to the first embodiment.

FIG. 3 is a diagram showing a display example of a result window according to the first embodiment.

FIG. 4 is a diagram showing a display example of a tile-type multi-window.

FIG. 5 is a flowchart showing a character recognition process according to the second embodiment.

[Figure 6]

[Figure 7]

FIG. 8 is a diagram showing how a polygon to be a recognition target area is drawn in an image window according to the second embodiment.

FIG. 9 is a diagram illustrating line segmentation by projection.

FIG. 10 is a diagram illustrating character segmentation by projection.

FIG. 11 is a diagram illustrating line and character cutout in a polygonal recognition target region according to the second embodiment.

[Fig. 12]

FIG. 13 is a diagram showing a manner of drawing a plurality of rectangles as recognition target areas in an image window according to the third embodiment.

FIG. 14 is a diagram illustrating line and character cutout in a polygonal recognition target area according to the third embodiment.

FIG. 15 is a flowchart showing a character recognition process according to a conventional example.

FIG. 16

FIG. 17

FIG. 18 is a diagram showing a display example of an image window according to a conventional example.

FIG. 19

FIG. 20

FIG. 21 is a diagram showing a display example of a result window according to a conventional example.

[Explanation of symbols]

 1 Scanner 2 Binarization Circuit 3 Mouse 4 Mouse Interface Circuit 5 CPU 6 ROM 7 RAM 8 Display 9 Display Interface Circuit 10 CPU Bus

Claims (7)

[Claims] 1. A character recognition method for recognizing a character drawn on an image original by reading the image original, the input step of reading the image original, and the image original read on the screen by the input step. A first display step of displaying in a first window; a designation step of designating an area in which character recognition is desired from the image original displayed in the first display step; and a designation step in the area designated by the designation step. A step of calculating the number of characters included in the area by cutting out the included characters one by one, and determining the size of the second window screen displaying the character recognition result based on the number of characters calculated by the calculating step. , A window opening step of opening the second window on the screen, and character recognition of characters included in the area one by one A character recognition method, comprising: a recognition step; and a second display step of displaying the recognition result of the recognition step in the second window. 2. The specifying step includes a rectangular specifying step of specifying a plurality of rectangles in the first window, and a plurality of rectangles are combined to form one polygon, and the polygon is used as the area. The character recognizing method according to claim 1, further comprising a polygon synthesizing step. 3. The step of designating comprises: drawing a plurality of line segments in the first window, forming one polygon by the plurality of line segments, and forming a region having the polygon as the region. The character recognition method according to claim 1, further comprising: 4. A character recognition device for reading an image original and recognizing characters drawn on the image original, comprising: input means for reading the image original; and the image original read by the input means on a screen. A first display unit for displaying in a first window; a designation unit for designating an area in which character recognition is desired to be performed from the image original displayed by the first display unit; and a region designated by the designation unit. Calculating the included characters one by one to calculate the number of characters included in the area, and determining the size of the second window screen displaying the character recognition result based on the number of characters calculated by the calculating unit. Window opening means for opening the second window on the screen, and character recognition of characters contained in the area one by one A character recognition device comprising: a recognition means and a second display means for displaying a recognition result by the recognition means in the second window. 5. The designating means composes a plurality of rectangles in the first window with a rectangle designating means for designating a plurality of rectangles, forms one polygon, and defines the polygon as the area. The character recognition device according to claim 4, further comprising a polygon synthesizing unit. 6. The designating means has a region forming means for drawing a plurality of line segments in the first window, forming one polygon by the plurality of line segments, and defining the polygon as the region. The character recognition device according to claim 4, wherein the character recognition device is a character recognition device. 7. The scanner, wherein the input means irradiates the image original with light, receives reflected light of the light and converts the reflected light into an analog electric signal, and converts the analog electric signal into a digital signal. 5. The character recognition device according to claim 4, further comprising a binarizing unit that outputs a binary signal.