JPH096892A - Information processing apparatus and method - Google Patents

Abstract

PURPOSE: To provide the information processor and its method which make it easy to decide quality and make obvious a character that should be noticed by controlling the output format of a 1st character candidate for a character image obtain by character recognition. CONSTITUTION: A document image set by an operator is read and inputted by an input part 2, and matches against standard patterns stored previously in a dictionary 9 for recognition to recognizes characters of the document image. When a recognition result is likely, the number of candidate characters become small and when the likelihood is small, more candidate characters are outputted. It is judged whether or not there are >=2 candidates obtained by the recognition processing for a character image of interest. When it is judged that there are >=2 candidates, the 1st candidate is displayed in red. When only one recognition candidate is obtained, on the other hand, the 1st candidate is displayed in blue. Consequently, a character having plural candidates and a character having only one candidate are quite obvious only looking at the display screen of the recognition results.

Description

Detailed Description of the Invention

[0001]

The present invention relates to an information processing apparatus and method,
More specifically, the present invention relates to an information processing apparatus and method for outputting a result obtained by character recognition.

[0002]

2. Description of the Related Art Generally, in a character recognition device, a document original is input as digital image data from an image scanner or the like, the distribution of characters is detected based on the input digital image data, and the character image is cut out for recognition processing. Take the procedure to do.

Then, on the display screen, a character string (sentence) composed of the first candidate in the candidate group obtained by the character recognition processing is displayed, and regarding the character designated by the operator, the second and subsequent candidates are displayed. It is necessary to display it, select one of them, and correct the sentence.

In some cases, there is no intended character in the candidates. In such a case, the correct character will be input from the keyboard or the like.

[0005]

However, in the above-mentioned conventional example, the recognized results must be thoroughly checked, and the burden on the operator cannot be ignored.

In the character recognition and correction process as described above, the character input operation is performed by another means such as a keyboard only after it is known that the intended character does not exist after the second candidate. In other words, useless work (work to confirm that there is no intended character after the second candidate)
Will force the operator.

[0007]

[Means for Solving the Problems] and

The present invention has been made in view of the above problems,
An object of the present invention is to provide an information processing apparatus and method that can easily determine the quality of a character recognition result as a whole and make it clear which character should be noted.

In order to solve this problem, for example, the information processing apparatus of the present invention has the following configuration. That is,
An information processing apparatus for outputting a result obtained by character recognition, comprising: an output unit for visually outputting a first candidate character of each character image obtained by character recognition; and a candidate for the obtained recognition result. Control means for controlling the output form of the first candidate according to a number.

According to a preferred embodiment of the present invention,
It is desirable that the control unit controls the output form of the first candidate depending on whether the number of candidates is n or more and when it is less than n.
As a result, it is possible to easily distinguish a character having a large number of candidates obtained by recognition and having a low probability and a character having a small number of candidates and a high probability.

Further, it is desirable that the threshold value n in the control means is "2". As a result, it becomes clear at a glance whether or not the number of candidates is only one, and it is also possible to know in advance whether or not the work of selecting a candidate will be wasted.

[0011]

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

FIG. 1 is a block diagram showing the schematic arrangement of a character recognition apparatus according to the embodiment. In FIG. 1, reference numeral 1 denotes the character recognition device of the first embodiment. Reference numeral 2 denotes an input unit for reading and inputting a document image, which is constituted by, for example, a scanner or the like, and converts the read original image data into digital and outputs it. Reference numeral 3 denotes a central processing unit (hereinafter referred to as CPU) for arithmetic processing that controls the entire apparatus 1. 4 is RO
M stores the control program and various data shown in the flowchart of FIG. Reference numeral 5 denotes a memory (RAM), which is used as a work area of the CPU 3 and stores the area for storing the document image data input from the input unit 2, the position and size of the character pattern cut out by the preprocessing unit 7. An area for storing the area to be stored is provided. A recognition result correction unit 6 corrects the character recognition result.
A preprocessing unit 7 cuts out a one-character pattern of the document image stored in the memory 5. A recognition unit 8 extracts geometrical features from the pattern preprocessed by the preprocessing unit 7,
The character recognition of the document image is performed by collating with the standard pattern stored in the recognition dictionary 9 in advance. A display unit 10 is an image display unit 1 for displaying an input document image.
4 and a recognition result display unit 15. Reference numeral 11 denotes a system bus, which includes a data bus from the CPU 3, an address bus, a control signal bus, and the like. Reference numeral 12 denotes an interface unit for performing interface control with an external output device such as the printer 13. Reference numeral 15 is a keyboard and a mouse (a kind of pointing device).

Generally, when performing character recognition processing, first, the position of each character image in the input image is detected and the character image is cut out. Then, a feature amount is extracted from the cut-out character image, the information is compared with the information stored in the recognition dictionary, and the character code having the closest feature amount is output as the first candidate and the second candidate.

When only one character candidate is output as a result of the character recognition processing, the certainty is judged to be high. On the contrary, the larger the number of candidates, the more difficult it is to differentiate the candidate group, and thus often means that the probability of the first candidate is low.

Also in this embodiment, description will be made assuming that such character recognition processing is adopted. The basis for determining that the probability of the first candidate is high is, for example, the probability of the first candidate and the second candidate obtained internally (the feature amount obtained from the character image of the recognition target, When the difference from the standard feature amount stored in the dictionary is small, the probability is high), the difference is more than a predetermined value. That is, this is the case where the first candidate has a sufficient likelihood difference with respect to the second candidate.

However, since the character recognition processing itself does not have the characteristics of the present invention, the description to this extent will be given here.
Details are omitted.

<Explanation of Recognition Result Window (FIG. 2)> FIG. 2 shows a window displayed on the screen of the result of recognition processing by the character recognition device 1 of the embodiment. Reference numeral 2 shown
Reference numeral 1 denotes a window which is an area for displaying the recognition result.

As shown in the figure, there are two types of display characters which are the recognition results in the embodiment. One is the reference numeral 22 shown
Is a red character and the other is a blue character indicated by reference numeral 23. Character 22 (character displayed in red) indicates that there are two or more recognized candidates.

When the operator designates a character displayed in red with a cursor (reference numeral 24 shown in the figure) interlocking with a mouse or the like, for example, the second candidate and below (the first candidate may be included)
Is displayed in the pop-up menu shown.

In the figure, the first candidate "OK" is also the selected character 1.
"Magazine", "Word" below the second candidate,
It shows that a total of 6 characters, "story,""tsume," and "complaint," were obtained as recognition candidates. If the number of candidate characters is more than a certain number, a scroll bar (a control for setting the scroll amount displayed at the right end or the bottom of the recognition result window 21 shown) is displayed in the pop-up menu.

As a result of the above, it is obvious to the operator as to whether or not the second candidate can be selected from the display screen. In other words, the higher the ratio of blue characters is, the better the recognition result of the entire sentence can be confirmed, and the character to be noted as the recognition result can be instantly determined.

FIG. 3 is a flowchart showing the processing contents of the above embodiment (procedures processed by the CPU 3), and the control program corresponding to this processing is stored in the ROM 4.

First, in step S31, the document image set by the operator is read and input by the input unit 2 (scanner) and stored in the memory 5.

Next, in step S32, geometric features are extracted from the individual character patterns preprocessed by the preprocessing unit 7 and collated with a standard pattern stored in the recognition dictionary 9 in advance to obtain a document image. Character recognition. When the recognition result is certain, the number of candidate characters is small, and when it is certain, many candidate characters are output.

Next, in step S33, the recognition result is displayed on the recognition result display section 14, and in step S34, the recognition result correction section 6 corrects the recognition result in accordance with the operator's instruction.

FIG. 4 is a flow chart showing the recognition result display processing of step S33 of FIG.

First, in step S41, the number of recognition result candidates for the target character image is checked. Here, there is one candidate for the character whose recognition result seems to be quite certain,
Or less. Incidentally, in the case of the recognition result “recognition” shown in FIG. 2, it indicates that there are a total of six recognition candidates in the pop-up menu.

Next, in step S42, it is determined whether the number of candidates obtained by the recognition process for the focused character image is two or more.

When it is determined that the number is 2 or more, the process proceeds to step S43, and the first candidate is displayed in red. On the other hand, when there is only one recognition candidate, the first candidate (1
Only) is displayed in blue.

Next, in step S45, it is determined whether or not the processing has been completed for all recognized characters. If not, the process returns to step S41 and the above processing is repeated.

As a result of the above, the character having a plurality of candidates and the character having only one candidate can be seen at a glance just by looking at the display screen of the recognition result. Therefore, it is possible to distinguish between a probable character and a suspicious character, and it is possible to distinguish a character that should be noted and a character that is not careful when checking the characters of a sentence.

Further, in the above example, the display screen is taken as an example of the output destination. However, when the printer 13 is adopted as the output destination, there are characters having candidates and characters having no candidates in the recognition result, and the recognition result Even when making corrections, they can be easily corrected without causing confusion. There are no or few probable characters, so by changing the attributes and displaying them, the user does not need to check all recognition results, but only the uncertain characters. It is possible to realize a character recognition device that is easy to use by the user, because the correction is considerably easy and it takes only a short time.

<Second Embodiment> In the above embodiment (first embodiment), the display form is changed depending on whether the threshold value of the number of candidate characters is 2 or more in step S42 of FIG.
The threshold n is not limited to “2” but “3” from the viewpoint that the credibility of character recognition as a whole can be judged when a sentence is displayed as the first candidate of the recognition result.
It may be above.

However, if the work of selecting the second candidate is to be performed, it is of course convenient to distinguish between the case of one recognition candidate and the case of two or more recognition candidates. Therefore, the user may freely set the threshold value.

<Third Embodiment> In the first embodiment, as shown in FIG.
In steps S43 and S44, the recognition result character is displayed in red when step S42 is satisfied, and is displayed in blue when step S42 is not satisfied. However, each color may be any color, and the color is not limited to a bold color. Characters and normal characters, with or without underline, font size (Mincho or Gothic, etc.) or font size, color depth, shaded, underlined, boxed or not, Any display form such as presence / absence, presence / absence of ruby, and whether or not it is a reverse character may be used. That is, each character may be displayed or output (including printout) in a different form depending on the number of candidates.

In some cases, the user may freely select the output form in the form of environment setting.

<Fourth Embodiment> In the first embodiment, as shown in FIG.
In step S41, the number of candidates is simply checked, but it is also possible to check only the number of candidates by character recognition, excluding characters input from a keyboard, that is, characters other than the character recognition result.

According to the above description, also in the fourth embodiment, it is possible to obtain the same operation and effect as those of the first embodiment.

<Fifth Embodiment> In the first embodiment, the size of the number of candidate characters is notified by any one of two types of display forms, but a plurality of stages may be used. For example, the number of candidate characters is 1, the number of candidate characters is 2 to m, and the number of candidate characters is m +.
1 to n, ... The output form may be output in three or more forms depending on the number of candidates, instead of the binary of the size of the candidates. Also in this case, if the operator can arbitrarily set (range of the number of candidates and output form), it becomes user-friendly.

According to the above description, also in the fifth embodiment, it is possible to obtain the same actions and effects as those in the first embodiment.

As described above, according to the present embodiment, there is a character having a candidate and a character having no candidate in the recognition result, and even if the recognition result is corrected, it can be easily corrected without causing confusion. Since there are no or few candidates for new characters, by changing the attributes and displaying them, the user does not need to check all recognition results, but rather should check the characters that are not certain, focusing on them. It is possible to reduce the burden on the operator due to the work.

It is needless to say that the present invention may be applied to a system composed of a single device or a plurality of devices, or may be realized by supplying a program to the device or system from the outside. Is. Therefore,
It is also possible to easily adapt to a storage medium having a program for realizing the present invention built therein.

In the embodiment, the document image to be processed has been described as being input from an image scanner or the like, but it may also be received when the image is stored on a storage medium (for example, a floppy disk) or a line. The invention of the present application is not limited to the above examples because it can be applied.

[0044]

As described above, according to the present invention, the quality of the character recognition result as a whole can be easily determined, and it becomes clear at a glance which character should be noted.

[0045]

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a character recognition device according to a first embodiment.

FIG. 2 is a diagram showing a display example of a recognition result in the character recognition device in the first embodiment.

FIG. 3 is a flowchart showing a process of the character recognition device in the first embodiment.

FIG. 4 is a flowchart showing details of step S33 in FIG.

[Explanation of symbols]

 1 character recognition device 2 input unit 3 CPU 4 ROM 5 memory 6 recognition result correction unit 7 preprocessing unit 8 recognition unit 9 recognition dictionary 10 display unit 11 system bus 12 interface unit 13 printer 14 image display unit 15 recognition result display unit 21 Recognition result window

Claims (6)

[Claims] 1. An information processing device for outputting a result obtained by character recognition, comprising: an output means for visually outputting a first candidate character of each character image obtained by character recognition; And a control unit that controls the output form of the first candidate according to the number of candidates of the recognition result. 2. The information processing apparatus according to claim 1, wherein the control unit controls the output form of the first candidate depending on whether the number of candidates is n or more or less than n. . 3. The information processing apparatus according to claim 2, wherein the threshold value n in the control means is “2”. 4. An information processing method for outputting a result obtained by character recognition, comprising: an output step of visually outputting a first candidate character of each character image obtained by character recognition; And a control step of controlling the output form of the first candidate according to the number of candidates of the recognition result. 5. The information processing method according to claim 4, wherein the control step controls the output form of the first candidate depending on whether the number of candidates is n or more or less than n. . 6. The information processing method according to claim 5, wherein the threshold value n in the control step is “2”.