JPH0820669B2 - Image information recording / reading method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention records an image information of a predetermined form and data on a microfilm, and an image for selecting and reading out necessary data from the microfilm. The present invention relates to an information recording / reading method.

(Technical background of the invention) In customer management cards, hospital patient cards, and the like, necessary data is entered on a card on which a form having a predetermined data entry field is printed. It is also widely practiced to save such a data-filled card as a microfilm. However, in the case of extracting necessary data from the microfilmed card, conventionally, the image of the microfilm is output to a reader or the like, read by human eyes and input to a keyboard or the like. For this reason, the work is difficult when reading data from a large number of cards. Also, if the card format is decided, it is possible to read only the predetermined position of this card with an OCR (optical character reader), but there is a problem that it can not cope when the card format changes to various kinds. there were.

(Object of the Invention) The present invention has been made in view of such circumstances,
Image information recording that allows the image data including the form and data to be copied onto the microfilm, while enabling the necessary data to be efficiently and easily extracted and read from the microfilm even if the form changes. -It is intended to provide a reading method.

(Structure of the Invention) According to the present invention, this object is recorded on a microfilm by combining an image of a predetermined form and an image of data imprinted in a predetermined area of the form, while it is required from the microfilm. In the image information recording / reading method of extracting and reading data, a form code for identifying the form is recorded on the microfilm, and at the time of reading, the form code corresponding to the form code is read by reading the form code. The image information characterized by reading the form from the form memory that stores the data, determining the position of the area in which the necessary data is imprinted, cutting out the image part of this data, encoding the data by image recognition and outputting. It is achieved by a recording / reading method.

(Operation) When recording a predetermined form and data on the microfilm, a form code for identifying the form is also recorded. When reading the desired data from the microfilm on which this form code is recorded,
First, the form code is read to identify the used form. The position (address) of the data to be read from this form is obtained, and the necessary data portion is selected and read from this position.

(Embodiment) FIG. 1 is a diagram showing a microfilm image of a customer management card used in an embodiment of the present invention, and FIG. 2 is a schematic diagram of an image information recording apparatus for recording the image of this card on a microfilm. FIG. 3 is a flow chart of the operation, and FIG. 4 is a schematic view of an image information reading device for reading an image on the microfilm.
FIG. 5 is a flow chart of the operation.

An image of a card as shown in FIG. 1 is recorded on the microfilm. In this figure, Form 10 has a number of columns separated by vertical and horizontal dividers, and odd columns (1, 3,
Lines 5 and 7) are item lines 12 in which item names are described.

The even-numbered rows are entry fields having entry fields in which data corresponding to each item above is entered. In this embodiment, a symbol indicating the type of the form 10, that is, the form code 16, is written on the upper left of the margin of the form 10 by the bar code display method. In addition, a check code for performing data check for each entry line 14 is provided on the right side of each entry line 14.
18 is marked by the bar code notation. In FIG. 1, the entered data is only numbers and katakana, but of course, names and addresses may be shown in kanji.

The card of FIG. 1 is recorded on microfilm by the recording device shown in FIG. In FIG. 2, 20
Is a laser light source such as an argon laser, 22 is an optical modulator, 24
Is a rotating polygon mirror, 26 is a lens, 28 is a galvanometer mirror, 30 is a lens, and 32 is a microfiche film. The light beam emitted from the light source 20 is on / off modulated by the light modulator 22, and then scanned in the main scanning direction (horizontal deflection direction) by the reflecting surface of the rotary polygon mirror 24. This rotating polygon mirror
The light beam reflected by 24 passes through a lens 26 and enters a galvanometer mirror 28. This galvanometer mirror 28
Deflects the light beam in the sub-scanning direction (vertical deflection direction) by one scanning line interval during one scanning by the rotary polygon mirror 24. As a result, the light beam that passes through the lens 30 and is converged on the film 32 is two-dimensionally scanned (raster scan) on the film 32.

The code and data of the form 10 recorded on the film 32 are designated by the keyboard 34. Of course, the data may be read from a memory (not shown) included in the computer (hereinafter referred to as CPU) 36 itself or from another CPU,
You may make it input directly from the keyboard 34.

The code designating the form 10 is entered into the form memory 38 that stores the patterns of the different forms 10 and the corresponding form 10 is selected. The form video signal of the corresponding form 10 is then output by the form generator 40.

On the other hand, the data designated by the keyboard 34 is transferred to the buffer memory 42 at a predetermined timing, and the character generator 44
At the data video signal. Also Form 10
Form code / video signal for displaying the form code 16 of FIG.
The check code video signal for displaying (see the figure) is output from the code generator 46. These form video signal, data video signal, form code
The video signal and the check code video signal are added by the adder circuit 48, and the driver 50 controls ON / OFF of the optical modulator 22 based on the added signal. The rotary polygon mirror 24, the galvanometer mirror 28, and the form, character, and code generators 40, 44, and 46 are controlled by the CPU 36, and the film 3
Each generator 4 corresponding to the scanning position of the light beam on 2
Video signals from 0, 44, and 46 are guided to the adder circuit 48 to control the optical modulator 22. As a result, as shown in FIG. 1, on the film 32, predetermined data is recorded in a predetermined data entry field of the form 10, and the form code 16 is also provided at a predetermined position.
And check code 18 are recorded.

The above-described image recording process is the procedure shown in FIG. That is, when the keyboard 34 designates the form 10 (step 100), the corresponding form 10 is read from the form memory 38 (step 102). Also, the data specified by the keyboard 34 is input from the keyboard 34 itself or CP
The data is read from the memory of U36, and the type of this data is input from the keyboard 34 (step 104). Since the address at which the data is recorded varies depending on the form 10, the address to be recorded is determined by the form 10 to be used (step 106). Then, the corresponding data, form or form code, and check code are respectively recorded at corresponding places (addresses) on the film 32 (step 108).

Next, an image information reading apparatus for extracting and reading out necessary data from the image information recorded on the microfiche film 32 in this way will be described with reference to FIGS.

In FIG. 4, 52 is a light source and 54 is a condenser lens, and the light of the light source 52 is guided to the microfiche film 32 by the condenser lens 54. The transmitted light of the film 32 is guided to the screen 60 by the projection lens 56 and the imaging lens 58, where it can be visually observed. Further, a part of the transmitted light of the film 32 is split by the half mirror 62, and this split light is imaged by the image sensor 66 by the imaging lens 64.
Image on top. As the image sensor 66, a one-dimensional line sensor that moves in a direction orthogonal to its length (sub-scanning direction), a two-dimensional area sensor, or the like can be used.
Since the output signal of the image sensor 66 includes the entire image of the card shown in FIG. 1, only the form code 16 indicating the type of the form 10 is read out from among them (see FIG. 5,
This form code 16 is entered into the form generator 68 (step 200). The form generator 68 reads the form 10 from the form memory 70 (may be common to the form memory 38 of the recording device) in which various forms are recorded (step 202), at which position on the card the data is recorded, In other words, the data address is obtained. The type of data you need in Form 10 is the keyboard
It is input from 72 (may be common to the keyboard 34 of the recording device) (step 204), and the address signal AD of this data is output to the data cutout circuit 76 described later (step 206). The form code 16 is recorded at a predetermined position on the card or at a fixed position on the form 10, and the form code 16 can be identified by reading the output of the image sensor 66 at the predetermined position.

An analog image output of the image sensor 66, that is, an image including the form 10 and data is input to the signal processing circuit 74 (step 208), where analog processing such as amplification, waveform shaping, and shading correction is performed (step 21).
0). The image signal processed in this way is output to the data cutting circuit 7
Although only the necessary data portion is cut out in step 6, the address of the required data obtained by the form generator 68 is based on the assumption that the size of the image input to the image sensor 66 is standardized to a constant value. And Therefore, this image signal is standardized to a certain size (step 212). In this case, it can be performed by detecting the diagonal corners of the form 10 or by detecting the lengths of two orthogonal sides of the form 10. Only the necessary data portion of the standardized image signal is cut out by the data cutting circuit 76 (step 214).

When only the data to be read out is extracted in this way, the image signal of this data is image-recognized according to a known image recognition process, the data on the microfilm is converted into a code of characters (including numbers), and the CPU is converted. And output to memory.

The image recognition process is performed as follows, for example.
First, the preprocessing circuit 78 cuts out the image signal for each character of the data, moves or rotates the center of the image as necessary, enlarges or reduces the image, normalizes it, and further binarizes it. (Step 216). At this time, character defects are repaired and, if necessary, thinning is performed to facilitate recognition.

When the preprocessing is performed in this way, the feature extraction circuit
At 80, feature extraction of image (text) is performed (step 21
8) Further, the recognition circuit 82 performs character recognition (step 2).
20). There are various methods of feature extraction. For example, paying attention to the features around the pattern, they are once roughly classified, divided into groups of similar characters, and the groups are finely distinguished to finally identify one character. A method is used.
Here, as a feature around the pattern, for example, a four-sided code method in which a long window is provided on each of four sides surrounding a character and the amount of a character line that enters the window is checked and coded is possible. It is also possible to use a peripheral feature based on a crossing distance from a frame circumscribing a kanji to a character line.

Characters can be classified according to horizontal and vertical complexity (complexity index), character stroke density, and stroke distribution method that uses the characteristics of the waveform of the intensity distribution of the projection of the character pattern on the coordinate axes. Can be used.

In this way, it is roughly classified by relatively low-order feature extraction,
After narrowing down to several tens to several hundreds of candidates, a character is determined by a method higher than this (step 82). For example, since Kanji is a combination of "hen" and "tsukuri", it is conceivable to use hierarchical structural analysis or tomographic matching in which pattern matching is repeated at each stage. In this method, in the first stage, a character pattern of 32x32 pixels is originally blurred by 4x4 pixels into 4-bit gradation,
The matching with the standard character dictionary 84 prepared separately is examined to narrow down the candidates. Next, as a second step, matching is similarly performed with a gradation of 8 × 8 pixels and 4 bits, and candidates are gradually narrowed down in this way.

It is also possible to use a method of quantifying the degree of deviation between the read character pattern and the standard character pattern of the dictionary using an evaluation scale such as "distance" or "similarity".

In any case, after the characteristics are extracted using various methods and the character recognition is completed, the post-processing circuit 86 performs post-processing (step 222). This uses the check code 18 read by the image sensor 66 to determine whether or not the correct character recognition is performed.
Check one line of 10 data entry lines 14. As a method of giving a check code, a known method such as a method of providing a check code for each byte of a character, a method of using a check code for each item, or a combination of several check codes can be used. As a check code,
For example, a parity check code whose total value is always an odd number can be adopted. If it is determined that there is an error in the character identification as a result of the check by the check code, the image identification processing is performed again by the re-identification circuit 88 using a more advanced image recognition method. For example, it is possible to use a method such as difference matching, which enables matching by emphasizing only a different portion from a similar character, or a method of using a stroke difference from a similar character (differential stroke determination).

In addition, when it is not possible to determine one character at the stage of character recognition, it is possible to select and determine the one having the highest matching probability from among the characters having the check code.

When the correct character identification is performed as described above, this data is coded and stored in the memory of the CPU (step 226). This data may be transferred to a CPU connected online or to another CPU offline so that they can be used mutually.

Although the recording / reading device is divided according to the functions in FIGS. 2 and 4 above, it is a matter of course that the CPU may have some or all of these functions.

Further, in the above-mentioned embodiment, only the position (address) at which the target data is present is obtained using the obtained form code. However, the present invention uses this form code for types such as Roman letters, numbers, etc., character size, font. The type and the like may be identified, and by doing so, the accuracy of image recognition can be further improved.

In the above embodiment, as shown in FIG. 1, form code 16
However, the present invention is not limited to this, and various ones such as a two-dimensional bar code display method can be adopted. As the form code, a notation that does not require complex pattern recognition such as characters is desirable, but a simple display of a predetermined font such as numbers can be used because it can be easily recognized.

(Effect of the Invention) As described above, the present invention records the form code for identifying the used form on the microfilm together with the image of the form and the image of the data, and at the time of reading the necessary data, this form is first used. To identify the form being used, determine where in the form the data is needed, extract and read the image portion of this data, and code this data by a method of image recognition. Since the necessary data is read by the method, the position (address) where the target data is can be easily obtained from the form code when extracting and reading the data from this form. Therefore, even if the form changes, the target data can be read easily and easily, and can be efficiently read.

[Brief description of drawings]

FIG. 1 is a diagram showing a microfilm image of a customer management card used in an embodiment of the present invention, and FIG. 2 is a schematic diagram of an image information recording apparatus for recording the image of this card on a microfilm, and FIG. Is a flow chart of the operation, FIG. 4 is a schematic view of an image information reading device for reading an image on the microfilm, and FIG. 5 is a flow chart of the operation. 10 …… Form, 16 …… Form code, 32 …… Microfilm, 70 …… Form memory.

Claims (1)

[Claims] 1. Image information in which an image of a predetermined form and an image of data imprinted in a predetermined area of the form are combined and recorded on a microfilm, and necessary data is extracted and read from the microfilm. In the recording / reading method, a form code for identifying the form is recorded on the microfilm, and at the time of reading, the form is read from the form memory for storing the form corresponding to the form code. An image information recording / reading method, which comprises: reading out, obtaining a position of an area in which necessary data is imprinted, cutting out an image portion of the data, encoding the data by reading image recognition, and outputting the encoded data.