JPH0581467A - Method and device for image processing - Google Patents

Abstract

PURPOSE:To provide the image processing method and device which process a read image by recognition, etc., to obtain processed data and perform correcting operation after the processing while displaying the original image and processed data. CONSTITUTION:Characters are segmented by processing read image data to obtain the images and positions of read characters as segmentation data, and the character images are recognized to obtain character codes as recognition data. Then the recognition data are displayed on the text screen of a CRT. The positions of the characters displayed on the text screen are made to correspond to the addresses of the corresponding recognition data and the both are combined together, so that the positions of the characters on the text screen and the character positions in the original image can be made to correspond to each other. (coordinate converting means). The screen is divided into a left and a right half part, which are a text screen and an image screen. On the image screen, a document image is displayed so that the left upper origin of the document is at the left upper part of the image when the image is read in; and the image on the specified text screen is processed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention performs processing such as recognition on an image read by an image scanner or the like, obtains processing data, displays an original image and processing data such as correction of recognition error, and performs processing. The present invention relates to an image processing method and apparatus that involve subsequent correction work.

[0002]

2. Description of the Related Art A character recognition device (OCR) will be described as an example. In OCR, a document is read by an image scanner or the like and the document image is displayed on a motor screen of a personal computer or the like. The area to be recognized in the document image is designated by the pointing device and characters are recognized, and the resulting character string is displayed as text on the monitor in the font that the personal computer has. After that, since the recognition result includes erroneous recognition, the calibration process is performed. It is easier to perform the proofreading process while comparing the text screen and the image screen displayed on the monitor than comparing with the paper of the original document.

FIG. 5 is an explanatory view of the prior art. J1
Is a 640 x 400 dot personal computer monitor screen, J2
Is a screen displaying a text or an image, J3 is a character of the recognition result displayed in the font of the personal computer,
In this example, it is schematically represented by □. J4 is a bit image displaying a character image of the original document image. In the example, switching between the text screen and the image screen is performed by a switch such as a function key. Further, the image screen cannot display the entire image (about 3400 × 4800 dots) when the A4 document is read at 400 dpi, so only a part of the image is displayed. To display the other part of the image, a scroll bar (not shown) is designated by a pointing device, or a screen scroll key assigned to the keyboard is pressed. On the text screen, the characters of the recognition result are displayed in a 16 × 16 dot font to compose the text, and editing operations such as insertion, deletion and replacement are possible. Also, the text screen can be scrolled like the image screen. Therefore, the operator performs the proofreading work and corrects the misrecognition while comparing the text and the image while switching between the text screen and the image screen. However, this method is inconvenient because you cannot see text and images at the same time. Also, as the proofreading work progresses, the display part of the text will be changed by scrolling, but in order to display the part of the original image corresponding to the character to be proofread, after switching to the image screen, Further, there is a great inconvenience in operation that the image has to be scrolled.

Therefore, as an improved version thereof, FIG.
An example is shown in. In the conventional example shown in FIG. 6, there is also a method of simultaneously displaying a text screen and an image screen, but this only eliminates the screen switching operation, and the inconvenience of the scroll operation remains.

As another conventional example, there is also a method of displaying one line together with a text / image screen as shown in FIG. 7 and displaying an image of the corresponding line of the text on the text. However, the inconvenience of horizontal scrolling still remains. In addition, it is usually necessary to proofread the text by looking at the entire sentence, not just at that line. For example, when almost the same character string appears in two or more places, the proofreading at such places is inconvenient in the present embodiment.

[0006]

As described above, in any of the conventional examples described above, in order to display the portion of the original image corresponding to the calibration portion, the operator must find the portion and scroll. There wasn't. Therefore, it is a case where processing such as recognition is performed on the original image, data after the processing is obtained, and calibration processing is performed on the processed data while comparing the original image and the processed data. It is desired to reduce the burden of calibration work by making it easier to perform.

[0007]

According to the present invention, image data is converted into processed data, and the image data and the processed data are displayed simultaneously or in a switching manner, and the position of the processed data and the position of the image data are displayed. The positions of the corresponding image data are identifiably displayed on the display screen when the positions are managed so as to correspond to each other and an arbitrary position of the processed data is designated.

[0008]

1 to 4 show a first embodiment of the present invention.

FIG. 1 is a block diagram of the present invention. 1 is a CPU for executing the processing procedure of the present invention, 2 is a bus, 3 is a CRT for displaying texts and images, 4 is a CPU and a CR.
T video interface, 5 VRAM for storing texts and images to be displayed on CRT, 6 image scanner for inputting image, 7 scanner interface, 8 pointing device for designating coordinates on CRT, 9 for pointing device The interface 10 is a program memory for storing the processing procedure of the CPU, 11 is a work RAM for storing a work area for the processing procedure and a document image, and 12 is a recognition unit for recognizing characters.

FIG. 2 is a flow chart showing the process of character recognition, and FIG. 3 is a flow chart showing the process of image interlocking. These processing steps are performed by the program memory 10
And is executed by the CPU 1. The steps will be described below in order.

In step S101, the image scanner 6
The document is read in and the binarized document image data D1 (image) is stored in the work RAM 11. In step S102, the document image data D1 is processed to cut out characters.
That is, first, the projection is performed in the horizontal direction (X-axis direction), and the character lines are divided by dividing at the projection breaks, then the projection is performed in the vertical direction (Y-axis direction) for each line, and the projection breaks are made. Characters are divided by dividing. In this way, for each character, the image can be cut out by the circumscribing rectangle, and the image can be divided for each character. The coordinates of the upper left point of the circumscribed rectangle of each character (X and Y coordinates when the upper left of the original document image is the origin is called the character position) are also obtained, and the cut character image and the character position are cut out data. D2 is set in the work RAM 11.

FIG. 8 is a diagram schematically showing the relationship between the cut-out character image and the character position. As is clear from the figure, the upper left of the original document image is set as the origin, and the character position is expressed as dot unit data (400, 800). It should be noted that this data is the dot output position, but it is sufficient if it corresponds to it, and it may be managed with a value of 1/128 thereof. In this way, the character position and the character image are stored in the work RAM 11 as the cut-out data D2. Further, as will be described later, the cutout data D2 and the recognition data D3 are further associated with each other and stored in the work RAM 11.

Next, in step S103, the character image is sent to the recognition section to recognize what kind of character it is and the character code as the recognition result is recognized as the recognition data D3.
Is put in the work RAM 11. The recognition algorithm uses a known method such as the direction index histogram method,
It need not be limited to this. At this time, the cut-out data D2 and the recognition data D3 are associated with each other in a one-to-one correspondence.
Next, in step S104, the recognition data is displayed on the text screen of the CRT as shown in FIG. Also at this time, it is associated with the address of the recognition data corresponding to the position of the character displayed on the text screen. By combining this correspondence with the former correspondence, it becomes possible to associate the character position on the text screen with the character position in the original original image. (Coordinate conversion means).
The screen of the CRT is divided into left half and right half as shown in Fig. 4,
It has a text screen and an image screen, respectively. The document image is displayed on the image screen so that the upper left origin of the document is located at the upper left of the image when the image is read.

When step S104 is completed, the apparatus is WA
It enters the IT state (step S201) and waits for coordinate input. Then, for example, "no" of the text screen of the example of FIG.
Character is specified by the pointing device.
Then, a pointing device interrupt occurs and W
AIT state is over. If the designated coordinates are read and the coordinates do not designate a character (determination in step S203), the process proceeds to step S202.
It returns to 01 and becomes a WAIT state. If the coordinates specify a character, the process proceeds to step S204. In step 204, which of the recognition data D3 the designated character corresponds to is reversed. And further cut data D
Find out which of the two corresponds. Since the cut-out data stores the coordinates (character position) of the character in the original image, the character position is set to be the image display base point coordinates. In step S205, the image is displayed again based on the new image display base point coordinates.

Then, as shown in FIG. 4, the state in which the original image corresponding to the character "NO" on the text screen is not displayed is changed to the display state in which the original image "NO" comes to the upper left of the image screen. ..

Next, in step S206, editing processing such as insertion, deletion, replacement, etc. is performed with respect to the designated character of the text screen.

Incidentally, in accordance with the editing process for the characters on the text screen, the corresponding image screen is also processed such as moving as a block of image data, and is inserted using the character pattern image on the text screen. Is also good.

Further, it goes without saying that the present invention may be achieved in an apparatus or system, or may be achieved by supplying a program to them.

(Other Embodiments) In the first embodiment, when a character is designated, the image is redisplayed so that the original image corresponding to the character is at the upper left of the image screen.
Alternatively, it may be at the center of the screen instead of at the upper left, or at any other predetermined position.

Further, although the first embodiment has been described with reference to horizontal writing, vertical writing is also possible.

The process of creating the processed data from the original image may be any process other than character recognition.

As described above, (1) According to the present invention, it is possible to efficiently perform the calibration work by comparing with the original image. (2) The original image does not necessarily have to be completely displayed on the monitor, and a partial display is sufficient. You can use an inexpensive monitor, and you can save memory. (3) Since the original image is enlarged and easy to see and can be compared with the processed data (text), the original image can be correctly captured.

[0023]

As described above in detail, according to the present invention, the image data is converted into the processed data, and the image data and the processed data are displayed simultaneously or in a switching manner, and the position of the processed data and the position of the image data are displayed. To manage the position,
When an arbitrary position of the processed data is designated, the position of the corresponding image data can be displayed on the display screen in a distinguishable manner.

[Brief description of drawings]

FIG. 1 is a block diagram of the present invention.

FIG. 2 is a flowchart of the present invention.

FIG. 3 is a flowchart of the present invention.

FIG. 4 is a diagram showing a display screen of the present invention.

FIG. 5 is a diagram showing a display screen of a conventional example.

FIG. 6 is a diagram showing a display screen of a conventional example.

FIG. 7 is a diagram showing a display screen of a conventional example.

FIG. 8 is a diagram showing a relationship between a character image and a character position.

[Explanation of symbols]

 1 CPU 2 Bus 3 CRT 10 Program memory 11 Work RAM 12 Recognition section J1 CRT screen J2 Text / image screen J3 Characters of recognition result J4 Original character image K1 Text screen K2 Image screen

Claims (8)

[Claims] 1. The image data is converted into processed data, the image data or the processed data is displayed, position information is managed so as to associate the position of the processed data with the position of the image data, and the desired position of the processed data is managed. An image processing method characterized in that a position of image data corresponding to a specified position is displayed on a display screen in a distinguishable manner. 2. A conversion means for converting image data into processed data, a display means for displaying the image data or the processed data, and a management means for managing position information so as to associate the position of the processed data with the position of the image data. Based on the position information managed by the management unit, the display unit displays the position of the corresponding image data on the display screen in a distinguishable manner when the desired position of the processed data is designated. Characteristic image processing device. 3. The image processing method according to claim 1, wherein the managed position information is related to an output dot position. 4. The image processing method according to claim 1, wherein the processed data is data after pattern recognition processing. 5. The image processing apparatus according to claim 2, wherein the managed positional information and the output dot position are related to each other. 6. The image processing apparatus according to claim 2, wherein the processing data obtained by the conversion means is data after pattern recognition processing. 7. The image processing apparatus according to claim 1, wherein the processed data is text data. 8. The image processing method according to claim 1, wherein the processed data is text data.