JPH0877290A - Character reader - Google Patents

Abstract

PURPOSE: To facilitate time management, operation planning, etc., for operation by an operator by calculating the read speed of documents and the number of sheets of documents in a unit time according to format information. CONSTITUTION: A process speed calculation part 7 scans a corresponding area on a document on the basis of the format information 720 downloaded from a host, i.e., information showing what position on the document, what kind of character, and how many characters are written at, and when the characters are read, the process time required for character reading (character recongition) is calculated from a specific calculation expression 722. The calculated time is displayed on the display screen of a display and correction part 4 and the operator can recognize the processing time for stacked documents. Further, the operator can be informed of the number of sheets processed in an inputted process time, and the process time required for the inputted number of documents on the basis of the process time and the number of sheets of documents inputted from the display and correction part 4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character reading device. Not only the character reading device but also any device can predict in advance to some extent how long it will take to carry out a certain work using the device. Is important in allocating resources and managing work.

Character reading devices, such as optical character reading devices (OCR), are often used in batches to process a large amount of slips (forms), and a large number of slips can be loaded and read at one time. To perform. At this time, to know how much amount and how much time can be processed,
It has a great meaning in the work performance of the operator.

[0003]

2. Description of the Related Art FIG. 7 is a diagram showing a configuration example of a conventional character reading device. First, from the character image on the form photoelectrically converted by the observation unit 1, the feature extraction unit 2 extracts the features required for character recognition. The extracted features are
In the next matching unit 3, the features of each category prepared in the dictionary 5 are matched in advance, and the most similar category is determined as the category of the input character.

Prior to reading the form, the above-mentioned character reading device writes format information regarding the form to be read from a host (not shown), that is, at which position of the form, how many characters of which character type are entered. The information indicating whether or not it has been downloaded is downloaded and stored in the format information section (disk file) 6 shown.

Therefore, every time the form is scanned, the format information of the corresponding area, that is, whether the character entered in the area (field) to be recognized is a numeral or a Chinese character. Information regarding the character type, etc. is given from the format information section 6, and which dictionary in the dictionary 5 to match is determined by this format information.

When the category is determined by the collation, the result is displayed on the display / correction unit 4 as a recognition result. If there is an error in the recognition result or there is a rejected character, the operator inputs correction information from the display / correction unit 4 and the corrected correct recognition result is returned to the host.

[0007]

Although the processing speed of the character reading device varies depending on the type of the form and the type of the entered character, the recognition time for each type of character and the paper feeding in the conveying section for conveying the form are performed. It can be theoretically calculated from the time.

However, this calculation formula is extremely complicated and it is not realistic to calculate it manually. In addition, if it is possible to predict the processing time based on the above formula and notify the operator, the man-machine interface (MM
Despite the fact that it is important from the above point of I), there is no conventional character reading device equipped with a mechanism for predicting the processing time and notifying the operator.

In view of the above-mentioned conventional drawbacks, the present invention predicts the processing time of the form in the character reading apparatus according to the amount of the stacked forms and notifies the operator of the processing time, or the processing time instructed by the operator. Then, based on the number of processed sheets, the collation range of the dictionary at the time of collation processing necessary for character recognition is dynamically set, and the processing time instructed by the operator,
An object of the present invention is to provide a character reading device that can complete the reading process within the number of processed sheets.

[0010]

FIG. 1 is a block diagram showing the principle of the present invention. The above problems can be solved by a character reading device configured as follows.

(1) The character recognition is performed by collating the character image obtained by the means for observing the characters described in the form or the characteristics obtained from the character image with a dictionary prepared in advance. A character reading device that performs reading, and that has format information corresponding to the form to be read and obtains information such as the reading position on the form, the number of characters, and the character type based on the format information A processing speed calculation unit for calculating the reading speed of the form or the number of sheets read in a unit time according to the format information is configured.

(2) When a predetermined number of sheets and a processing time are input from the display / correction unit provided in the character reading device, the ratio information to the normal reading time from the processing speed calculation means and the above Based on the format information, the collation range control unit that sets the collation area of the dictionary is configured so that the reading of the designated number of documents is completed within the designated processing time.

[0013]

In the character reading apparatus of the present invention, based on the format information downloaded from the host, that is, the information indicating at which position of the form and what kind of character is written, the corresponding area of the form. When a character is scanned by scanning, a processing speed calculation unit 7 that calculates a processing time required for character reading (character recognition) based on a predetermined calculation formula is provided, and the processing speed calculation unit is provided. 7
By displaying the processing time calculated in step 3 on the display screen of the display / correction unit 4, the operator can recognize the processing time of the form that has just been loaded. In addition, based on the processing time and the number of processed sheets input from the display / correction unit 4, notify the operator of the number of processed sheets within the input processing time and the processing time required to process the input number of documents. This makes it easier to manage time and perform work plans for performing work such as appropriate response to the next work.

Further, the character reading apparatus of the present invention has a dictionary for collation at the time of character recognition based on the processing speed e obtained from the processing speed calculation unit 7 and the processing time e input by the operator. A collation range control unit 8 is provided which calculates the capacity and determines the collation range based on this capacity, the character type and the number of characters f from the above format information.

Generally, in character recognition processing, when the collation capacity of the dictionary increases, the recognition time also increases, but the recognition rate is improved. Therefore, in the character reading apparatus of the present invention, by providing the collation range control unit 8, the operator can voluntarily prioritize the processing time and the recognition rate, depending on the content of the work and the operator's time margin. It is possible to select whether or not to read.

Therefore, by knowing the relationship between the amount of processing for reading the form and the processing time, the operator can easily perform time management, work planning and the like when performing work. or,
Depending on the situation of character reading, such as when the operator has time to spare or when the form cannot be misread, the capacity of the dictionary to be matched is changed, that is, the recognition rate is changed. Appropriate characters can be read.

[0017]

Embodiments of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a diagram showing the principle of the present invention.
FIG. 6 is a diagram showing an embodiment of the present invention.
Shows a configuration example of a processing speed calculation unit, FIG. 3 shows a configuration example of a matching range control unit, FIG. 4 shows a reading speed calculation example, and FIG. 5 shows a processing outline of the matching range control unit. A flow chart is shown, and FIG. 6 shows an example of the structure of the dictionary.

In the present invention, the character image obtained by the means for observing the characters written on the form, or the characteristic obtained from the character image is collated with a dictionary prepared in advance, thereby A character reading device that performs recognition, and that has format information corresponding to the form to be read, and that reads by obtaining information such as the reading position, the number of characters, and the character type on the form from the format information. Therefore, according to the format information, a processing speed calculation means for calculating the reading speed of the form or the number of sheets read in a unit time, and a predetermined number of forms from the display / correction unit provided in the character reading device When the processing time is entered,
Based on the ratio information to the normal reading time from the processing speed calculation means and the format information,
The collation range control means for setting the collation area of the dictionary so that the reading of the designated number of forms is completed within the designated processing time is a means necessary for carrying out the present invention. . The same reference numerals indicate the same objects throughout the drawings.

The structure and operation of the character reading apparatus of the present invention will be described below with reference to FIGS. Also in the character reading apparatus of the present invention, the basic operation of character recognition does not differ from the operation described with reference to FIG. 7 and thus is omitted, and the processing speed calculation unit shown in FIG.
A configuration example of the collation range control unit 8 and its operation will be mainly described.

First, FIG. 2 shows an example of the configuration of the processing speed calculation unit explained in FIG. 1, which is a microprocessor (MPU).
The processing speed is calculated by the 70 executing a predetermined control program in the control memory 71. At this time, the format information expanded in the format information area 720 in the main memory (MM) 72 and the constant table 721a, 721b
Is referenced. Since the constant tables 721a and 721b differ depending on the paper feeding direction of the form, they generally correspond to the reading direction of the form, that is, the direction of the character string, or the direction different from the character string direction by 90 degrees. Thus, two types of constant tables 721a and 721b are prepared.

As described above, the reading speed of the form varies depending on the reading direction of the form, the character reading device, etc., but in general, the speed of the transport system, the size of the form, the field position, the number of characters, and It is determined by the recognition speed of each character.

FIG. 4 (a) shows an example of a form. For example, when the length of the first form in the feeding direction is L mm,
The processing speed of one form is calculated by the calculation formula shown in the general OCR manual as shown in FIG. 4 (b). In this calculation formula, C1: a constant unique to the character reading device C2: a constant related to the speed of the transport system C3: a constant related to the reading speed of the character C4: a constant related to the entry rate of the form C5: a constant related to the field of the form Is. The above constants correspond to the reading direction of the form, as shown in FIG.
It is prepared as 21a and 721b.

The microprocessor (MPU) 70 executes a predetermined control program to calculate the processing speed per one sheet based on the calculation formula shown in FIG. 4 (b). At this time, the format information transmitted from the host (information indicating which character type and N characters are written at which position) and the constant tables 721a and 721b are referred to. {Corresponding to the invention described in claim 1} The reading speed T of the form calculated based on the equation of FIG. 4 (b) above, or the number of sheets read per unit time
T) is displayed on the display / correction unit 4 shown in FIG. 1 before the reading of the form is started. {Corresponding to the invention described in claim 2} Next, at the time of operation, depending on the character entry rate on the form,
In the example of recalculating the processing speed, the constant C4 is corrected according to the character entry rate of the real form. In this case, an appropriate default value is empirically set as an initial value, and the processing speed calculated before the start of reading is calculated using this default value. {Corresponding to the invention described in claim 3} In this way, the average reading speed of the forms read up to that point or the number of sheets read in a unit time is calculated, and further, each time one sheet is read. By displaying the recalculated value in a certain area of the display screen of the display / correction unit 4, for example, a predetermined area at the upper right, the processing speed can be presented to the operator every moment.

In the above embodiment, the processing speed for the form is calculated based on the format information transferred from the host (not shown) based on the instruction from the operator. When the operator inputs a predetermined processing time from the correction unit (input unit) 4, the input time is calculated from the reading speed T shown in FIG. 4 (b) and the time T'input by the operator. The number of documents that can be read in T'is, for example,
It may be calculated by (time / T) × T ′ and displayed on the display / correction unit 4. {Corresponding to the invention described in claim 4} Further, when the time T'instructed by the operator has passed, the reading of the forms is automatically completed, and the number of the forms read by that time is determined as the processing speed. It may be stored in the main storage device (MM) 72 of the calculation unit 7 and displayed on the display screen of the display / correction unit 4 based on an instruction from the operator.

When the operator instructs the number of sheets to be read, the time required for reading the number is displayed in a predetermined area of the display / correction unit 4 so that the operator can continue this work. You can decide whether or not to do it.

Next, an embodiment corresponding to claims 5 and 6 will be described. FIG. 3 shows an example of the configuration of the matching range control unit 8 of FIG. 1, in which a format information area 820 for storing the format information transferred from the host is stored on the main storage device (MM) 82. Each character type of the processing speed storage area 821 that stores the processing speed T calculated by the processing speed calculation unit 7, the reading time T2 input by the operator, and the dictionary 5 that is referred to when character recognition is performed. A collation start position pointer table 823 for storing the start address of the stored area and a collation end position pointer table 822 for storing the collation end position pointer corresponding to each character type are provided.

In the formula for calculating the processing speed T shown in FIG. 4B, C3 is a constant relating to the reading speed for reading each character, as described above. The reading speed of individual characters is slow if the dictionary 5 to be matched has a large capacity. Therefore, at the processing time instructed by the operator,
When the processing speed is changed, the constant C3 is newly obtained and the collation range of the dictionary 5 is determined from the new constant C3.

The processing in this case will be described with reference to the flowchart of FIG. First, the normal reading speed (processing speed) T is calculated based on the calculation formula shown in FIG. 4 (b).
Here, from the display / correction unit 4, by the operator, for example, the reading time per sheet (time / number of documents) T2
Is entered. {Refer to processing step 100 in FIG. 5} The matching range control unit 8 calculates T2 / T = k, and uses the calculated k to calculate the processing speed shown in FIG. 4 (b). Substituting into equation T, the corresponding constant C3 'is calculated as shown in process step 102. {Refer to processing steps 101 and 102 in Fig. 5} Now, assuming that the capacity of the dictionary 5 at the time of standard processing is DNold and the capacity of the dictionary 5 at the processing time T2 is DN, the calculation formula DNold * C3 'shown in processing step 103 is given. The capacity DN of the dictionary 5 corresponding to the processing time T2 can be obtained by / C3 → DN. {Refer to processing step 103 in FIG. 5} FIGS. 6 (a) and 6 (b) schematically show a configuration example of a dictionary in the conventional example and a dictionary to which the present invention is applied. In the conventional example shown in FIG. 6 (a), if the character type (numeric, kanji, ...) is simply known from the format information, the start address of the area storing the character type is shown in FIG. 3, for example. The collation start position pointer table 823 is used for instructing, and the collation process with the dictionary 5 for character recognition is performed from the address indicated by the collation start position pointer table 823. Then, the end of the collation has a collation end mark (indicated by ▪) in the dictionary 5, and the collation process is ended by detecting this end mark.

As described above, when the processing speed is changed within the processing time designated by the operator, the constant C3 is set to C.
3'is converted so that only the corresponding matching area is matched. Specifically, as shown in Fig. 6 (b),
The dictionary 5 has an added dictionary in addition to the standard dictionary, and when the collation range is expanded with the calculation of the above constants, the collation range control unit 8 of the present invention sets the collation end mark, From the Y position (indicated by ■) in FIG. 6 (b),
When it is necessary to rewrite to the position of Z (indicated by □) which is the collation range of the additional dictionary or to reduce the collation range, FIG.
Rewrite to the position of X in (b) (shown by ◯).
{Refer to processing step 104 of FIG. 5} In the embodiment shown in FIG. 3, the collation end position pointer address is previously stored in the main storage device (MM) 82 of the collation range control unit 8 in correspondence with the value of k. Are stored in a table format, and when k is obtained in the process of calculating k, the matching end position pointer table 822 is referenced to recognize the corresponding matching end position pointer address, and the matching end position pointer address is recognized. The process ends. Of course, in the process of obtaining k, the matching end position pointer address may be calculated in real time. Alternatively, a predetermined end code may be set at the corresponding position and the end code may be detected in the matching process.

In this way, when the number of sheets and the time designated by the operator are input, the collation range of the dictionary 5 is referred to by the collation end position pointer 822 so that the reading process is completed within that time. To adjust.

Therefore, when there is a time margin for the character reading process according to the operator's instruction, the recognition rate can be improved by increasing the collation capacity of the dictionary. {Examples corresponding to the invention described in claims 8 and 9} As described above, the character reading device of the present invention is a character image obtained by a means for observing a character described in a form,
Alternatively, a character reading device for recognizing a character by collating a feature obtained from a character image with a dictionary prepared in advance, and having format information corresponding to the form to be read, and its format A character reading device for reading by obtaining information such as the reading position on the form, the number of characters, and the character type based on the information, and calculating the reading speed of the form or the number of sheets read in a unit time based on the format information. Means for doing so. Another feature is that the collation range of the above-mentioned dictionary is set so that the reading is completed within the number of sheets and the processing time by inputting the number of sheets and the processing time designated by the user.

[0032]

As described above in detail, according to the present invention, by knowing the relationship between the processing amount and the processing time of the form, the operator can manage the time and the work plan when performing the work. Easier to do. Also, depending on the status of the form processing, such as when there is enough time or when the form cannot be misread, the capacity of the collation dictionary is changed, that is, the recognition rate is changed to change the situation. The corresponding character can be read.

[Brief description of drawings]

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

FIG. 2 is a diagram showing an embodiment of the present invention (No. 1).

FIG. 3 is a diagram showing an embodiment of the present invention (part 2).

FIG. 4 shows an embodiment of the present invention (part 3).

FIG. 5 is a diagram showing an embodiment of the present invention (No. 4).

FIG. 6 is a diagram showing an embodiment of the present invention (No. 5).

FIG. 7 is a diagram showing a configuration example of a conventional character reading device.

[Explanation of symbols]

1 Observation unit 2 Feature extraction unit 3 Collation unit 4 Display / correction unit 5 Dictionary 6 Format information unit 7 Processing speed calculation unit 70 Microprocessor (MPU) 71 Control memory 72 Main memory (MM) 720 Format information area 721a, b Constant Table 722 Processing speed calculation formula 8 Collation range control unit 80 Microprocessor (MPU) 81 Control memory 82 Main memory (MM) 820 Format information area 821 Processing speed storage area 822 Collation end position pointer table 823 Collation start position pointer table 100 to 104 Processing step X, Y, Z Collation end position

Claims (6)

[Claims] 1. A character reading device which has format information corresponding to the form such as a reading position on the form to be read, the number of characters, and a character type, and which reads the form according to the format information, based on the format information, A character reading device comprising a processing speed calculation means for calculating a reading speed of the form. 2. The character reading device according to claim 1, further comprising a display unit for displaying information obtained by calculating the reading speed of the form by the processing speed calculating means. 3. The character reading apparatus according to claim 1, wherein the processing speed calculation means reads the number of sheets, and then the reading speed of the sheet based on the number of characters actually entered in the sheet. A character reading device characterized in that 4. The character reading apparatus according to claim 1, wherein an input section for inputting a predetermined processing designated time and a form which can be read within the designated time calculated by the processing speed calculation means. A character reading device having a display unit for displaying the number of sheets. 5. The character reading device according to claim 1, wherein an input unit for inputting a predetermined number of sheets and a reading of the specified number of sheets calculated by the processing speed calculation means. A character reading device having a display unit for displaying a required time. 6. The character reading device according to claim 1, wherein an input unit for inputting a predetermined number of sheets and a processing time, ratio information to a normal reading time from the processing speed calculation means, and A collation range control unit that sets a collation area of the dictionary so that the reading of the designated number of forms is completed within the designated processing time based on the format information. Character reading device.