JPH05324902A - Image input method - Google Patents

Abstract

(57) [Summary] [Purpose] When inputting an image, it is possible to perform efficient work without having to confirm the orientation of the read original image on the display. [Structure] First, in step 21, the counter 4 is initialized to 1, and the counters 5 and 6 are initialized to 0. In step 22, a character is cut out from the original image, in step 23, the cut-out character is recognized, and if the recognition is successful in step 24, the counter 5 is incremented by 1 in step 25. If the recognition is not successful, the cut-out character is rotated 180 ° in step 26,
The character rotated by 180 ° is subjected to recognition processing in step 27, and if recognition is successful in step 28, the counter 6 is incremented by 1 in step 29. When these processes are completed, 1 is added to the counter 4 in step 31. The above process is repeated 6 times, and finally the direction in which the number of times of successful recognition is large is the correct direction in step 32, and if it is the opposite direction, the original image is rotated 180 ° in step 33.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image input method used in an image filing system for registering a document image read by a scanner in a recording medium such as an optical disk.

[0002]

2. Description of the Related Art Conventionally, in an image filing system, a document consisting of a plurality of pages can be automatically and continuously read, and then a keyword can be added and registered in a recording medium. When this function is used, the worker is relieved from the document registration work while reading a plurality of pages, so that the work efficiency can be improved. However, when a document having a plurality of pages is registered in this way, some pages in the document may be reversed in direction, and as it is, only that page is registered in reverse direction. In order to prevent such a situation, conventionally, the direction of the document is checked on the display every time one sheet is read, or after all pages have been read once, one page is displayed on the display to display each page. The orientation of the page was checked, and if there was an opposite orientation, the image was rotated 180 ° each time.

[0003]

As described above, in the above-described conventional image input method, even if a plurality of pages are automatically and continuously read, the operator will end up with the orientation of the image on the display over the entire page. Therefore, there is a problem that it is necessary to confirm the above, and improvement of work efficiency is required.

The present invention solves the above problems, and an object of the present invention is to provide an image input method which enables efficient work without the need to confirm the orientation of the read original image on the display. It is a thing.

[0005]

In order to solve the above-mentioned problems, the image input method of the present invention includes a first step of recognizing character data cut out from a read original image, and the first step. A second step of determining the orientation of the original image based on the result of character recognition in step 3, and a third step of rotating the read original image by 180 ° based on the determination result in the second step; A fourth step of recognizing the character data rotated by 180 ° in step 3 and a fifth step of determining the orientation of the character data rotated by 180 ° based on the character recognition result in the fourth step. Steps,
The sixth step in which these determinations are repeated a plurality of times for other cut out character data and the number of successful recognitions in the second step and the fourth step are compared, and based on the result, the read original is read. The configuration has a seventh step of rotating the image by 180 °.

[0006]

With the above structure, the orientation of the original image is determined by recognizing the character image cut out from the read original image, and the read character image is rotated 180 ° and rotated 180 ° according to the determination result. The number of successful recognitions before the rotation is compared with the number of successful recognitions after the 180 ° rotation, and it is determined whether or not the entire original image is rotated with the direction having the largest number of successful recognitions as the correct direction.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of an image filing system using an image input method according to an embodiment of the present invention. In FIG. 1, reference numeral 1 denotes a central processing unit (hereinafter referred to as CPU) that controls the system and performs data processing,
2 is a read-only memory (hereinafter R
OM), and the CPU 1 controls the system and processes data according to the programs stored in the ROM 2. 3 is a random access memory (hereinafter referred to as RAM), 4, 5 and 6 are counters, 7 is a recognized character storage area for storing a character image cut out from an original image for character recognition, 8 is an image memory, Reference numeral 9 is an external storage device such as an optical disk, 10 is an input unit such as a keyboard or mouse, 11 is a CRT display device (hereinafter referred to as CRT),
Reference numeral 12 is an image scanner for reading a document, and 13 is a printer.

The operation of the image filing system thus configured will be described below with reference to the flowchart of FIG. First, a document image is optically read by the image scanner 12 (step 1) and written in the image memory 8 as binarized image data (step 2). Next, the orientation of the original image in the image memory 8 is automatically determined and rotated by 180 ° if necessary (step 3), and the process proceeds to the next page (step 4).

The processing from step 1 to step 4 is
The process is repeated until all pages of the document to be registered are read. As a result, a plurality of pages of image data are stored in the image memory 8 as shown in FIG. The image data of the read document is stored in the external storage device 9 together with the keyword (step 5) input from the input unit 10 (step 6). The image data stored in the external storage device 9 is searched for by the keyword input from the input unit 10, and C
It is output to the RT 11 and the printer 13.

The process in step 3 described above will be described in detail below with reference to the flowchart of FIG. First, the counter 4 is initialized to 1, and the counters 5 and 6 are initialized to 0 (step 21). When initialization is complete, the original image stored in the image memory 8
Image data of one character is cut out and recognized character storage area 7
(Step 22). Next, the character image data in the recognized character storage area 7 is subjected to character recognition processing (step 23), the value of the counter 5 is incremented by 1 when the recognition is successful (steps 24 and 25), and when the recognition fails. The character image data in the recognized character storage area 7 is 180
It is rotated (steps 24 and 26). Then 180
The rotated character image data is subjected to recognition processing (step 27), and if the recognition is successful, the counter 6 is incremented by 1 (steps 28 and 29).

Here, the extraction of a character image from image data will be described. As shown in FIG. 5, the image data in the image memory 8 is logically divided into six blocks, and the character closest to the upper left corner is cut out for each block. The processing from step 22 to step 30 is repeated for the characters in the original image while incrementing the counter 4 by 1 each time (step 31). As a result, the counter 5 stores the number of successful recognitions of the character image data before rotation, and the counter 6 stores the number of successful recognitions of the character image data after rotation.

When the processing for 6 characters is completed, the values of the counter 5 and the counter 6 are compared (step 30), and if the value of the counter 6 is larger, the image data in the image memory 8 is rotated by 180 °. (Step 33). In other words, character recognition is performed for each character before and after 180 ° rotation, and the one with the most successful recognition is determined to be the correct orientation. If the number of recognitions after the rotation is large, the read image data is Rotate 180 ° in the image memory 8.

As described above, in the present embodiment, the character data is cut out from the document image read from the image scanner 12 and recognized to automatically determine the orientation of the document, and the page read in the reverse direction is detected. , Automatically 180 °
Since the image is inverted and recorded in the external storage device 9, at the time of document registration, the operator only has to set the original document and register the keyword after the reading is completed, which enables efficient and efficient document registration work. ..

[0014]

As described above, according to the present invention, the character data cut out from the read original image is character-recognized, the direction of the character data is judged based on the character recognition result, and this judgment is performed. Based on the result, the character data is rotated by 180 °, and the character data rotated by 180 ° is subjected to character recognition.
The direction of the character data rotated by 0 ° is discriminated, and the discrimination is repeated a plurality of times including other character data cut out, and the number of successful recognitions when rotated 180 ° and when not rotated are compared. Since the original image read based on is rotated by 180 °, even if the original to be registered includes an original set in the opposite direction,
It is automatically inverted and registered. Therefore, it is not necessary for the worker to check the orientation of the image read while looking at the display as in the conventional case, and efficient registration work can be performed without any trouble.

[Brief description of drawings]

FIG. 1 is a block diagram of an electronic filing system using an image input method according to an embodiment of the present invention.

FIG. 2 is a flowchart of an image input method according to an embodiment of the present invention.

FIG. 3 is a detailed flowchart of a main part of the image input method according to the embodiment of the present invention.

FIG. 4 is a diagram illustrating a storage state of image data in the image input method according to the embodiment of the present invention.

FIG. 5 is a diagram illustrating character segmentation in the image input method according to the embodiment of the present invention.

[Explanation of symbols]

 1 CPU 2 ROM 3 RAM 8 Image Memory 9 External Storage Device 10 Input Unit 11 CRT 12 Image Scanner 13 Printer

Claims (1)

[Claims] 1. When recording a document image read by an image scanner on a recording medium such as an optical disc, a first step of recognizing character data cut out from the read document image, and a character in the first step. Based on the recognition result,
A second step of determining the orientation of the character data, a third step of rotating the character data by 180 ° based on the determination result of the second step, and a 180 ° rotation of the third step. A fourth step of recognizing the character data obtained by the character recognition, and a character recognition result in the fourth step,
A fifth step of determining the orientation of the character data by rotating it by 180 °, a sixth step of repeating these determinations a plurality of times including other cut out character data, the second step and the fifth step And a seventh step of rotating the read original image by 180 ° based on the result of comparison of the number of successful recognitions in the step.