JPH10326305A - Time recording device, and computer-readable recording medium where time recording process program is recorded - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enhance the efficiecy of an input operation for a fingerprint image and a fingerprint matching process. SOLUTION: A fingerprint image obtained by reading a fingerprint of an examinee, an item of recording time, and the fingerprint read time are inputted from a data input part 2 to a 1st matching process part 10. The 1st matching process part 10 compares the input data with registered data for matching which are registered in a staff master file 20a in advance to match the fingerprint and specify the examinee, and clocking-on/off data are recorded in a clocking-on/off data file 20b. If a matching error occurs, the input data in this case are inputted to a 2nd matching process part 40, matching operation is performed again to specify the examinee, and the clocking-on/off data of the examinee are recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a time recording apparatus for managing working hours and the like on the basis of the time at which an employee leaves or leaves the company.

[0002]

2. Description of the Related Art Conventionally, a time card has been prepared for each employee, and when an employee leaves or leaves the company, his / her time card is printed and recorded on a time stamp machine using a time stamp machine. It is widely practiced to collect and manage the working hours of employees.

[0003]

In such a work time management of the time card system, the work time manager needs to collect time cards, and it is difficult to automate the work by a machine.

[0004] In addition, the recording work on the time card is left to each employee, and it is not possible for the person performing the recording operation to recognize the individual, such as whether he / she is using his / her own time card. There is a problem that the time is recorded on another person's time card, and that an illegal time recording or forgery is possible.

On the other hand, a fingerprint collation device for recognizing an individual by collating a fingerprint image obtained by reading a fingerprint with a pre-registered fingerprint image has already been put into practical use. The introduction of a system that automates time management is being considered.

However, in a fingerprint collation device,
Even if the collation performance is high, it is inevitable that a collation error such as a failure in collation due to the degree of contamination, scratches, drying, etc. of the finger, despite the registered fingerprint.

When a collation error occurs, the fingerprint image is input again and the collation operation is executed again, so that the collation operation takes time. Therefore, if such a fingerprint recording time recording system is introduced into factories or the like where many employees come and go at the same time, if the fingerprint verification time is prolonged due to a verification error, a queue of employees waiting for fingerprint input will be created. In the worst case, there is a risk that the opening hours will be delayed.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a time recording apparatus capable of performing a fingerprint input operation smoothly and performing highly reliable time recording and management. Make it an issue.

[0009]

Means for Solving the Problems In order to solve the above problems,
The time recording device of the present invention includes a reading unit that reads a fingerprint of a subject to obtain a fingerprint image, a setting unit that sets items for time management, and a clock unit that measures the date and time when the fingerprint was read. Information input means for generating input information from a fingerprint image, management item information, and fingerprint reading date and time information obtained from each means; personal registration information having a registered fingerprint image stored in advance in a storage means; Matching means for identifying the subject by comparing the input information with the input information, and recording the management item information and the date and time information input from the information input means in a time management data file of the subject identified by the matching means. And a time recording device, wherein the matching means has a main matching means and a sub-matching means, and the main matching means has the information input means. When a matching error occurs because personal registration information corresponding to the input information from the stage cannot be detected, the matching operation on the input information is abandoned, while the sub-matching unit abandons the matching operation by the main matching unit. The input information is collated again (corresponding to claim 1).

According to such a configuration, even if a collation error occurs in the main collation processing unit, it is not necessary to re-enter the fingerprint image, the collation time required for one subject can be reduced, and the fingerprint image input operation by the subject can be shortened. Can be performed smoothly.
Also, for the fingerprint image in which the collation error has occurred, the sub-collation processing means performs the collation processing again, identifies the subject, and records the date and time information, so that highly reliable time recording and management can be performed for each subject. it can.

The main collation unit performs a collation operation for a predetermined time from a predetermined time when the frequency of information input by the information input unit increases, while the sub collation unit operates the main collation unit. It is preferable that the collation operation is performed when not performed (corresponding to claim 2). According to such a configuration, the main collation unit can operate in response to a case where the frequency of information input by the information input unit is high and the collation in a short period is required, When the means is not operating, that is, the frequency of information input is low, and the comparison operation is performed for all input information that resulted in a verification error in a time zone where the time required for verification is relatively long,
The efficiency of the fingerprint image input operation and the matching process can be improved without impairing the matching accuracy.

[0012] The main collating means collates the personal registration information based on code information obtained by performing predetermined image processing on a fingerprint image of the input information and encoding the fingerprint information. When the collation is performed, the fingerprint image input from the information input unit is recorded as a reference image in the collated personal registration information, and the sub-collation unit includes:
It is preferable that the reference image and the input fingerprint image are compared to perform a collation operation. (Corresponding to claim 3).

According to such a configuration, the input operation of the fingerprint image can be performed more smoothly, and the collation operation can be surely performed even for the non-collated data. The main collation unit stores the input information that could not be collated as the uncollated information in the storage unit and performs a collation operation on the next input information, while the sub collation unit performs the collation on the uncollated information. It is preferable that a collation operation is performed, and the uncollated information that has been collated is deleted from the storage means (corresponding to claim 4).

According to such a configuration, the uncollated information is temporarily stored in the storage means, so that the main collation means and the sub-collation means can be independently executed without being interfered with each other's operations. Since the collation operation can be performed, the efficiency of the collation processing in each collation unit can be improved.

[0015] It is preferable that the sub-collating means be configured to repeatedly execute the collation operation a plurality of times when a collation error occurs (corresponding to claim 5). According to such a configuration, even if the input information has a collation error in the main collation unit, there is a high possibility that the input information is collated by the sub collation unit, and time recording by the recording unit can be performed smoothly. .

Further, it is desirable that the sub-collating means execute the collating operation while gradually reducing the collating accuracy.
(Corresponding to claim 6). According to such a configuration, the possibility of collation by the sub-collation unit can be further increased, and time recording by the recording unit can be performed more smoothly.

Further, there is provided a notifying means for notifying the end of the collation operation, and the main collation means notifies the completion of the collation operation irrespective of the collation result corresponding to the input information from the information input means. It is desirable that the means be operated so as to notify the test subject (corresponding to claim 7).

According to such a configuration, the subject can recognize the completion of the collation operation, so that the next subject can be promptly prompted for a fingerprint input operation, and the efficiency of the fingerprint image input operation can be improved.

Further, the time recording apparatus of the present invention comprises: reading means for reading a fingerprint of a subject to obtain a fingerprint image; setting means for setting time management items; and time measuring means for measuring the date and time when the fingerprint was read. Information input means for generating input information from a fingerprint image, management item information, and fingerprint date and time information obtained from each of these means; and personal registration information having a registered fingerprint image stored in advance in storage means. A collation unit for collating the input information from the information input unit to identify the subject; and a management item information and date and time input from the information input unit in the time management data file of the subject identified by the collation unit. A time recording device having recording means for recording information, wherein the collating means have different collating accuracy, and are connected in ascending order of collating accuracy. A plurality of processing means having a hierarchical structure in which the personal registration information corresponding to the input information cannot be detected. Aside from abandoning, the input information is transferred to a lower-level processing unit whose matching accuracy is lower than its own matching accuracy, and a matching operation for the next input information is executed (corresponding to claim 8).

According to such a configuration, the processing means having different collation accuracy perform the collation operations independently and in parallel, so that the collation can be performed at a high speed and even if a collation error occurs, a plurality of collations can be performed. Since the collating operation is performed by the collating means, the input operation of the fingerprint image by the subject can be performed smoothly, and highly reliable time recording and management can be performed.

Further, the recording means records the item information set by the setting means and the date and time information by the time keeping means in the data file for time management of the subject identified by the collating means, and the personal registration information. While the item information and the fingerprint image at the time of collation are recorded, the collation unit, when the collation is performed, separates the item information recorded in the personal record information and the item information of the input information inputted this time. It is determined whether or not the personal record information has been previously collated by the lower-level processing means. If it is determined that the personal record information has been collated by the lower-level processing means, It is preferable that the time management data file and the personal registration information be recorded based on the above.

According to such a configuration, the recording means always records information in accordance with the collation result having the highest collation accuracy, so that the collation operation can be speeded up and the reliability of the recorded information can be improved. Can be compatible.

Further, when the matching means matches the personal registration information corresponding to the input information, the matching means attaches an identification flag to the personal registration information indicating that the matching has been completed. It is desirable to create a subject list that has not yet been collated based on the presence or absence of the flag, and to cause the recording means to record the subject list (corresponding to claim 10).

According to such a configuration, it is possible to identify a subject who has performed a fingerprint input that cannot be verified by the verification means based on the unverified subject list. Inconvenience such as not being performed can be prevented.

Further, the present invention provides a first step of reading a fingerprint of a subject to obtain a fingerprint image, a second step of setting time management items, and a third step of measuring the date and time when the fingerprint was read. A fourth step of identifying the subject by comparing personal registration information having a registered fingerprint image stored in advance in the storage means with the fingerprint image, management item information, and date and time information of fingerprint reading; A fifth step of recording the input management item information and the date and time information in the time management data file; and, if the subject cannot be identified in the fourth step, the collation accuracy is reduced and the subject is identified again. A computer-readable recording medium in which a program for causing a computer to execute the six steps is recorded (corresponding to claim 11).

The recording medium is a CD-RO.
M (COMPACT DISK READ ONLY MEMORY), MO (MAGNET
OPTIC DISK).

[0027]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a time recording apparatus according to the present invention will be described below.

FIG. 1 is a block diagram schematically showing a time recording apparatus according to a first embodiment of the present invention. The data input unit 2 includes an imaging unit 2a that images the finger of the subject employee,
It comprises an operation unit 2b in which button switches (not shown) operated by the examinee are arranged, and a clock unit 2c for measuring date and time, and corresponds to a fingerprint image by the imaging unit 2a and button operation of the operation unit 2b. The three items of the input button number data and the input date and time at which the image was input are output.

That is, the imaging unit 2a, the operation unit 2b, and the clock unit 2C correspond to reading means, setting means, and timekeeping means, respectively, and the data input unit 2 corresponds to information input means. A button switch (not shown) of the operation unit 2b is used to designate an item for time management.
"Leave", "Leave a break", and "Return to a break" are prepared, and the button number data value is 1,
2, 3, and 4 are assigned.

The operation unit 2b (button switch)
Is integrated with the imaging unit 2a, that is, by providing the same number of imaging units 2a as the number of items and pressing the imaging unit 2a corresponding to the item selected by the subject, the reading of the fingerprint and the setting of the item can be performed simultaneously. Can also. In addition, during the time set by the administrator, a predetermined item may be selected without the subject operating the operation unit 2b.

The first collation processing unit 10 is composed of, for example, a microcomputer, and performs predetermined image processing on a fingerprint image from input data input from the data input unit 2 to extract a characteristic portion of the fingerprint. Based on a predetermined collation algorithm, data to be collated is generated based on registered collation data (search key) of a reference data file (hereinafter referred to as an employee master file) registered in advance in the external recording device 20. Then, the registered collation data that matches the collation data is searched to identify the subject employee. First
As a result of the search, if there is registered data that matches the data to be verified in the employee master file 20a (in the case of a search HIT), the matching processing unit 10 stores each of the data stored in the external storage device 20 as the external storage means. The employee's attendance data is read out from the employee's attendance data file 20b, and is updated and recorded based on the button number data and the input time from the data input unit 2.

On the other hand, as a result of the search, if there is no registered data that matches the data to be verified (in the case of search NG), the data to be verified is stored as unverified data in the non-verified data file 20C of the external storage device 20. .

Further, the first collation processing unit 10 displays the start of the collation operation, the completion of the collation operation, and the like on the display unit 40, and notifies the start of the collation operation, the completion of the collation operation, and the like to the examinee. The display unit 30 is constituted by, for example, a liquid crystal display panel, and notifies the examinee from the first collation processing unit 10 of the collation operation status and the like.

That is, the first collation processing unit 10 corresponds to a main collation unit and a recording unit, the display unit 20 corresponds to a display unit, and the employee master file 20a corresponds to the personal registration information.
The attendance data files 20b correspond to time management data files, respectively.

The second collation processing unit 40 is composed of, for example, a microcomputer. The uncollated data generated by the first collation processing unit 10 is directly used without using the coded data to be collated this time. Then, a comparison process is performed by comparing the fingerprint images.

The collation operation is repeatedly performed with respect to the input data that is the collation NG while gradually lowering the collation accuracy, and the uncollated data that has been collated is deleted from the external storage device 20.

That is, the second collation processing unit 40 corresponds to a sub collation unit and a recording unit. Then, the first collation processing unit 10 is configured to provide two hours from 7:00 am to 9:00 am, two hours from 11:00 am to 1:00 pm, for example, two hours from 7:00 am to 9:00 am It operates for two hours from 5:00 pm to 7:00 pm.

On the other hand, the second collation processor 40 is configured to operate when the first collation processor 10 is not operating. The first collation processing unit 10 performs a retrieval process based on the data to be collated coded for collation, so that high collation accuracy and high-speed retrieval are possible.
This is advantageous for a collation operation under a situation where the frequency of collation of fingerprints becomes high, such as during lunch break.

However, as described above, depending on the state of the fingerprint, correct collation may not be possible. on the other hand,
In the second collation processing unit 40, since the fingerprint image is compared with each of a plurality of images registered in advance to perform collation, the collation time is longer than that of the first collation processing unit 10, but the first collation processing unit Even if the fingerprint image 20 fails to be verified, the verification is often successful. For this reason, the operation time of the first collation processing unit 10 and the second collation processing unit 40 is set as described above,
The overall efficiency of the matching process is improved without impairing the matching accuracy.

The first collation processor 10 and the second collation processor 40 are separated for convenience of explanation.
Both may be constituted by a single microcomputer. The external storage device 20 is composed of, for example, a hard disk drive, and has the above-described employee master file 20.
a, an attendance data file 20b and an uncollated data file 20c are stored, and are written and read by the first collation processor 10 and the second collation processor 40.

The data items of the employee master file 20a are registration collation data (search key), employee identification number, and input data of the last N times (fingerprint image, button number data, and input time). ), Operation unit 2
There are provided five items, namely, a flag (hereinafter, referred to as a mode flag) corresponding to each button switch (not shown) of b, and a collated flag set when the search HIT or collation is OK.

In this embodiment, the button switch is 4
Since there are types, the mode flag is prepared for 4 bits. The first collation processing unit 10 and the second collation processing unit 40
At the time of search HIT or collation OK, the collated flag is set to 1, the operated button switch is grasped from the value of the button number data, and the bit of the mode flag corresponding to this is set to 1. ing.

The mode flag of past data stored in the employee master file 20a always has a different bit flag from the mode flag of previous input data. For example, it is impossible for the same person to go to work twice in a day without leaving the company.

Therefore, the mode flag is used in the first collation processing unit 10 to determine an operation error of the button switch of the subject employee based on the state of the mode flag of the past data.

The data for the above five items is assigned to each employee as one record, and a set of these records constitutes the employee master file 20a. The attendance data file 20b includes three items: time, employee identification number, and button number data.

The data items of the uncollated file are the same as the input data described above, and are three items of a fingerprint image, button number data, and input time. FIG. 2 is a flowchart illustrating the operation of the first matching processing unit 10. The first collation processing unit 10 uses a built-in clock (not shown) to set the current time to the work start time (7:00 am) and the lunch break start time (11 am
Hour), and when it is time to leave the company (5 pm), the processing of this program is repeatedly executed for 2 hours.

That is, the first collation processing unit 10 performs collation processing for a predetermined time from a predetermined time when the frequency of information input by the information input means increases. First collation processing unit 10
Monitors the presence or absence of data input from the data input unit 2,
It waits for data input (step 101). When data is input from the data input unit 2, image processing is performed on a fingerprint image of the input data, and a characteristic portion of the fingerprint is extracted and coded data for verification is created (step 102). . Next, the registered matching data (search key) in the employee master file 20a that matches the data to be matched is searched using a predetermined matching algorithm (steps 105 and S106).

As a result of the search, if there is matching registered collation data (in the case of the search HIT), the employee identification number is grasped from the record of the registered collation data, and the employee identification number corresponding to the employee identification number from the attendance data file 20b is obtained. The system reads out the attendance file and writes and records the input time and button number (steps S110 and S111).

Next, returning to the employee master file 20a again, the collated flag indicating that the corresponding record has been collated is set to 1, and the mode flag bit corresponding to the button number data is set to 1 and The input data (fingerprint image, input time, button number data) input this time is overwritten on the area where the oldest input data is stored in the area where the past data is stored, and the contents of the past data are updated. Is displayed on the display unit 30 to notify that the fingerprint collation has been completed for a certain period of time (steps S112, S113, and S114).

On the other hand, as a result of the search in step S106, if there is no registered data that matches the data to be verified and there is a verification error (in the case of search NG), the data to be verified is later processed in the second verification processing. The uncollated data file 2 in the external storage device 20 is used as the uncollated data used by the unit 40.
0C, and a display for notifying that the fingerprint matching has been completed for a certain period of time is performed on the display unit 30 (Step S).
130, S114).

Then, it is determined whether or not the current time has reached a predetermined operation end time. If not, the flow returns to step S101 to continue the above operation, while if the operation end time has been reached, Then, the program ends (step S140). In the present embodiment, the operation end time is, for example, 9:00 am, 1:00 pm, and 7:00 pm.

That is, if the first collation processing unit 10 fails to recognize the employee as a result of the fingerprint collation as a result of the fingerprint collation, the first collation processing unit 10 leaves the re-collation processing to the second processing unit 40, and The employee is notified of the completion of the fingerprint collation operation without requesting the re-input of the fingerprint image, and waits for the input of the next subject. Thus, the examinee can pass through the data input unit 2 without delay if the fingerprint input is performed only once, and the efficiency of the fingerprint input operation can be improved.

FIG. 3 is a flowchart showing the operation of the second collation processing unit 40. The second collation processing unit 40 performs collation again using the input data that could not be collated by the first collation processing unit 10, and updates the time and attendance file.

First, it is monitored whether or not the first collation processing unit 10 is operating, and waits until the operation of the first collation processing unit 10 stops (step S201). This is because, as described above, the processing in the first matching processing unit 10 is prioritized due to the processing time, and the processing in the second matching processing unit 40 is executed outside business hours when fingerprint input is not performed. This is to improve the efficiency of the fingerprint collation and the recording process of the attendance file.

Next, uncollated data is read from the uncollated data file 20C to obtain an image to be collated, button number data, and the like (step S202). Then, from the employee master file 20a, individual fingerprint images registered in the past data are sequentially read as registered images for records in which the verified flag is not set (the verified flag is 0), and the predetermined algorithm is used. Based on the search, an image that matches the image to be verified is searched (step 205, S20).
6).

As a result of the search, when a matching registered fingerprint image exists (in the case of collation OK), the employee identification number is grasped from the record including the registered image, and the employee identification number is registered in the attendance data file 20b. The file is read, and the input time and the button number are written and recorded (steps S210, S211).

Next, the collated flag of the record is set to 1
And sets the mode flag bit corresponding to the button number data to 1, and sets the input data (fingerprint image, input time, button number data) to the oldest input data in the area where the past data is stored. Is overwritten to update the content of the input data for N times (steps S212 and S213).

Then, the input data read in step S202 is deleted from the uncollated data file 20C (step S214). On the other hand, as a result of the search in step 206, if there is no registered data that matches the data to be verified and a verification error has occurred (in the case of verification NG),
It is determined whether or not the current matching accuracy is a predetermined minimum accuracy (step S220).

Note that the collation accuracy here means, for example,
The following description is based on the assumption that the number of coincidences of fingerprint feature points used in the fingerprint image collation algorithm is higher, and that the greater the number of coincidences, the higher the collation accuracy.

If the matching accuracy has not reached the minimum accuracy, the matching accuracy is reduced by one step (the number of matching of the feature points is reduced by a predetermined number) in step S230, and the operation from step 205 is executed again. Then, the search operation is repeatedly executed until the matching accuracy becomes minimum.

On the other hand, if the collation precision has reached the minimum precision, the collated data is stored again as uncollated data in the uncollated data file 20C of the external storage device 20, while the collated flag is not set. All uncollated records are extracted, list data of uncollated employees is created, and stored in the external storage device 20 as uncollated employee files (steps S222 and S240).

Then, when the process proceeds to step 245 via step S214 or step S240, all the uncollated data recorded in the external storage device 20 is read.
It is determined whether or not the collation processing has been performed. If uncollated data remains, the process returns to step S202 to repeat the collation processing. If the collation processing has been performed on all uncollated data, step S112 (see FIG. 2) and step S21
All the collated flags set in step 2 are reset (step S247), and the process ends in preparation for the next collation operation by the first collation processing unit 10.

The finally remaining uncollated data file 20
C and the uncollated employee list data are read into, for example, a printer (not shown) and printed on a recording sheet as an uncollated employee list, and used for confirmation by an administrator.

That is, in the second collation processing section 40, the uncollated data file 20C of the external storage device 20 is stored.
The non-collating data stored in the employee master file 20a is read out, and the fingerprint image therein is compared with the registered image in the past data stored in the employee master file 20a to recognize the employee and manage the working hours of the employee. The first collation processing unit 1
Unlike 0, when the collation cannot be performed, the collation operation is repeatedly performed while gradually lowering the collation accuracy, and the uncollated data is digested, so that the efficiency of the collation processing can be improved.

In the case of the first embodiment, the second matching processing section 40
Reprocesses the input data that could not be processed by the first matching processing unit 10 after the operation of the first matching processing unit 10 is completed.
This is a mode in which a collation error recovery process in the collation processing unit 10 is performed, and is suitable when the work start time and the work end time of the employee are fixed. However, when the present invention is applied to a work place or the like in which shift work in which the start and end times of work are shifted is introduced, the following second embodiment in which a plurality of matching processing units always operate in parallel is desirable.

FIG. 4 is a block diagram showing a schematic configuration of a time recording apparatus according to the second embodiment. The same components as those in the first embodiment are denoted by the same reference numerals, and description thereof is omitted.

In the case of the second embodiment, the first collation processing unit 10
0, the second matching processing unit 400, a third matching processing unit 500 having a lower matching accuracy than the second matching processing unit 400, and a fourth matching processing unit 600 having a lower matching accuracy than the second matching processing unit 400. Is provided.

That is, the first to fourth matching processing units 100 to 600 correspond to matching means and recording means. For the input data from the data input unit 2, the collation operation by the first collation processing unit 100 is given top priority,
The matching processing unit 400 determines whether a matching error has occurred in the first processing unit 100, the third matching processing unit 500 determines whether a matching error has occurred in the second matching processing unit, and the fourth matching processing unit has a matching error. Each operation is performed when a collation error occurs in the 3 collation processing unit 500. That is, when a matching error occurs in the upper processing unit, the lower processing unit starts.

A plurality of third collation processing units 500 are provided, and input data having a collation error from the second collation processing unit 400 is passed to each third collation processing unit 500 one by one. It has become.

Similarly, a plurality of fourth collation processing units 600 are provided, and input data having a collation error in each of the third collation processing units 500 are input, and a collation operation with further reduced collation accuracy is executed. Is done.

The data items of the employee master file 200a include the registered collation data (search key), the employee identification number, and the input data of the last N times (fingerprint image, button number data, input time) in the first embodiment. In addition to the flags corresponding to the button switches (not shown) of the operation unit 2b, five items are newly prepared to which new values indicating the collation accuracy (another exclusion rate) have been newly added N times. The data for the items is assigned to each employee as one record.

FIG. 5 is a flow chart showing the operation of the first collation processing section 100. Unless the power of the time recording device is turned off, the first collation processing section 100 will be described with reference to FIG.
Is repeatedly executed.

The first collation processing section 100 includes a data input section 2
(Step S301), similar to the first embodiment, extracts the characteristic part of the fingerprint of the fingerprint image input from the data input unit 2 and encodes the data to be verified and the registered verification data (search key). ), Based on a predetermined collation algorithm, the registered collation data matching the collated data is searched from the employee master file 200a (steps S302 to S306).

If no matching registered collation data is found in step S306 and a collation error occurs, the input data is transferred to the second collation processing unit 400 (queuing), and the display unit 30 displays the data for a certain period of time. Then, a display notifying that the fingerprint collation has been completed is performed, and the process ends (steps S330 and S314).

In step S 306, if there is matching registration data (at the time of search HIT), the data input unit 2
It is determined whether or not the button number data input from the user and the latest button number data in the past data in the employee master file 200a match (step S).
309, step S310).

In this embodiment, a plurality of collation processing units operate in parallel, and the employee master file 200a
And the time and attendance file 200b are updated and recorded. For this reason, when they match, it can be determined that the previous collation operation was performed by the lower collation processing unit. Therefore, in this case, the first matching processing unit 100 having high matching accuracy is used.
Is performed again, and the process proceeds to step S320 and subsequent steps in order to further increase the reliability of the comparison result.

First, the latest input data from the past data in the employee master file 200a is recorded in another recording area, temporarily saved (step S320), and the input data from the data input section 2 is used. Then, the latest input data of the past data is updated (step S321).

Here, there are cases where the attendance data recorded with the collation OK in the lower collation processor (hereinafter, referred to as pre-collation) has low reliability. For this reason, the time and attendance data of the pre-verification is invalidated by setting an invalid flag that regards the data as invalid (step S322), and the latest time and attendance data is newly generated based on the current verification result. Recording is performed (step S323).

Then, the pre-collation is regarded as an erroneous judgment, and the saved input data (input data at the pre-collation) is transferred to the second collation processing unit 400 and processed again (step S324). This is because the result of the pre-collation processing is canceled by the operations of steps S321 to S323.

Then, a display for notifying that the fingerprint collation has been completed is performed on the display unit 30 for a certain period of time (step S).
314). On the other hand, if the button number data does not match the latest button number data in the past data, it can be recognized as the first collation operation, so that step S311 is performed.
The process proceeds from step S313 to step S313 to record the time and attendance data in the time and attendance data file 200b,
The current input data is written and updated in the past data of the batch (the same processing as steps S111 to S113 in FIG. 2).

Then, the display unit 30 is caused to perform a display for notifying that the fingerprint collation has been completed for a certain period of time (step S3).
14), end the process. As described above, the first collation processing unit 100 determines whether or not the pre-collation has been performed when the collation is successful, and again performs the first collation processing unit when the pre-collation has been performed. 100 to rewrite and update the time and attendance file data based on the verification in
Erroneous recording of time and attendance file data due to erroneous verification in the lower-level verification processing unit is prevented.

If the collation fails, the second
The input data is passed to the matching processing unit 400, and the matching of the input data is entrusted to the second matching processing unit 400. For this reason, even when a collation error occurs, the collation operation is performed by another collation unit, so that it is not necessary to request the examinee to input a fingerprint again. The efficiency of the recognition operation of the employee can be improved together.

FIG. 6 is a flowchart showing the operation of the second collation processing unit 400. Unless the power of the time recording device is turned off, the second collation processing unit 400 operates as shown in FIG.
Is repeatedly executed. First, the first matching processing unit 1
The process waits until the input data is transferred (queued) from 00 (step S401). When the transferred input data is received (step S402), the first record of the employee master file 200a is read (step S403). If the read record is not the end of the file (step S404), it is determined whether or not the button number data of the input data matches the latest button number data in the past data recorded in the employee master file 200a (step S404). Step S405).

It should be noted that, at step S404, EOF
When (END OF FILE) is recognized, the input data and the matching accuracy value are transferred as parameters to the third matching processing unit 500 waiting for data input among the plurality of third matching processing units 500 ( Step S430).

If the button number data matches in step S405, it is determined whether or not the current collation accuracy value (collation accuracy value in the second collation processing unit 400) is higher than the collation accuracy value of the past data. (Step S40)
6) If the matching accuracy value is high, it is determined whether or not the fingerprint image of the input data matches the registered image of the past data (step S407).
Move to 0.

If the button number data does not match in step S405, it can be determined that the pre-collation has not been performed, so that the process proceeds to step S407 without comparing the collation accuracy values.

If the current collation accuracy value is less than or equal to the past data collation accuracy value in step S406, and if the fingerprint images do not match (if collation error occurs) in step S407, The next record in the employee master file 200a is read (step S44).
1), becomes a loop routine of step S404. The reason why the comparison accuracy values are compared in step S406 is to prevent the collation processing from being performed on the records that have been collated by the first collation processing unit 100 (upper-level collation processing unit).

If the fingerprint images match in step S407, it is again determined whether or not the button number data of the input data matches the latest button number data in the past data in the employee master file 200a. (Step S410).

If both match, the latest input data from the past data is recorded in another recording area, saved once, and the input data received from the first collation processing unit 100 is saved. To update the latest input data of the past data and to compare the matching accuracy value with the second matching processing unit 400.
Is updated to a value corresponding to the collation accuracy (step S42).
0, step S421).

Next, with respect to the time and attendance data obtained by the pre-collation, the data is invalidated by setting an invalid flag which regards the data as invalid (step S422), and the latest time and attendance data is newly updated based on the current collation result. Generate and record (step S423).

Then, the pre-collation is regarded as an erroneous judgment, and the saved collation data (the input data at the pre-collation) and the collation accuracy value are used as parameters for the third collation processor 50.
0 to be processed again (step S424).

On the other hand, if the button number data does not match the latest button number data in the past data in step S410, there is no need to update the time and attendance data again by the pre-collation, so that the time and attendance data is recorded. ,
The process of setting the collated flag, updating past data for the past N times, and updating the collation accuracy value is completed (steps S411, S412, S413).

As described above, the second collation processing unit 40
In the case of 0, the collation operation is performed again on the queued input data having a collation error, and even if the collation succeeds, if the pre-collation has been performed, the second collation processing unit 400 And rewrites and updates the time and attendance file data based on the comparison in. In addition, the input data for which the matching result in the lower-order matching processing unit is found to be incorrect by the processing of the second matching processing unit 400 is transferred to the third matching processing unit 500.

Although the same processing is performed in the first collation processing unit 100, compared with the case where such processing is performed only by the first collation processing unit 100, the time of attendance due to erroneous collation is reduced from an earlier stage. There is an advantage that erroneous recording of file data can be prevented.

FIG. 7 is a flow chart showing the operation of the third collation processing section 500. Unless the power of the time recording device is turned off, the third collation processing section 500 is executed.
Is repeatedly executed.

The third collation processing unit 500 performs the collation operation with a lower collation accuracy than the second collation processing unit 400 and performs the collation operation on a single input data. The same operation as that of the two matching processing unit 400 is performed.

Waiting for data to be input from second collation processing section 400 (step S501), and receiving the transferred input data (step S502), reads out the first record of employee master file 200a (step S503). If the read record is not the end of the file (step S504), it is determined whether the button number data of the input data matches the latest button number data in the past data recorded in the employee master file 200a. (Step S50
5).

It should be noted that, at the step S504, the EOF
When (END OF FILE) is recognized, the input data and the matching accuracy value are transferred as parameters to the fourth matching processing unit 500 waiting for data input among the plurality of fourth matching processing units 600 ( Step S530).

If the button number data matches in step S505, it is determined whether or not the current collation accuracy value (collation accuracy value in the third collation processor 500) is higher than the past data collation accuracy value. (Step S50)
6) If the matching accuracy value is high, it is determined whether or not the fingerprint image of the input data matches the registered image of the past data (step S507).
Move to 0.

If the button number data does not match in step S505, it can be determined that the pre-collation has not been performed, so that the process proceeds to step S507 without comparing the collation accuracy values.

If the current collation accuracy value is less than or equal to the past data collation accuracy value in step S506, and if the fingerprint images do not match (in the case of a collation error) in step S507, The next record in the employee master file 200a is read (step S54).
1) A loop routine of step S504 is performed. If the fingerprint images match in step S507, it is determined whether or not the latest matching accuracy is a predetermined minimum accuracy corresponding to the matching accuracy in the fourth matching processing unit 600 (step S507).
510).

If the collation accuracy is the lowest value, pre-collation has been performed by the fourth collation processing unit 600 at the lowest order. Rewrite and update the past data and the matching accuracy value (steps S521, S522, S5
23).

Then, the pre-collation is regarded as an erroneous judgment, and the fourth collation processing unit 60 uses the saved input data (input data at the pre-collation) and the collation accuracy value as parameters.
0 to be processed again (step S524).

If the collation accuracy is not the minimum accuracy, it is regarded that there is no pre-collation, and the record of time and attendance data, the setting of a collated flag, the updating of past data for the past N times, and the updating of the collation accuracy value are performed. (Step S51)
1, S512, S513).

As described above, the third collation processing section 50
Even in the case of 0, even if the collation is OK, if the pre-collation has been performed, the time and attendance file data is rewritten and updated based on the collation operation in the third collation processing unit 500. Also, the input data for which the third collation processing unit 500 has determined that the collation result in the lower collation processing unit is incorrect is transferred to the fourth collation processing unit 600.

As described above, even if the user enters any one of the first to third collation processing sections, the clock data file is always updated in real time based on the collation based on the collation accuracy of the collation processing section. File data is always maintained based on the matching process with the highest matching accuracy,
It is possible to improve the efficiency of fingerprint input work for the examinee and also to improve the reliability of employee recognition.

FIG. 8 is a flowchart showing the operation of the fourth collation processing section 600. The fourth collation processing section 600 repeatedly executes the processing of FIG. 8 unless the power of the time recording device is turned off.

The process waits until data is input from the third collation processing unit 500 (step S601). When the transferred input data is received (step S602), the first record of the employee master file 200a is read (step S603). If the read record is not the end of the file (step S604), it is determined whether or not the button number data of the input data matches the latest button number data in the past data recorded in the employee master file 200a. (Step S60
5). If the button number data does not match, the fingerprint image of the input data is compared with the registered image of the past data in the employee master file 200a at the predetermined lowest matching accuracy, and whether or not the images match is determined. It is determined (step S607).

At the step S604, EOF (END OF
FILE) is recognized, there is no collation processing unit lower than the fourth collation processing unit 600 in the present embodiment.
The input data is stored in the external storage device 20 as uncollated data that cannot be collated anymore, and the process is terminated (step S6).
22). If the button number data matches in step S605, the pre-collation is performed by the fourth collation processing unit 600.
Are recorded in the non-collating data because the collation processing is performed by the higher-order collation processing unit including the data (step S6).
22). When the pre-collation has been performed by the fourth collation processing unit 600, both the pre-collation data and the current collation data may be treated as uncollated data.

If the fingerprint images do not match in step S607, the next record in the employee master file 200a is read (step S641), and the loop routine of step S604 is performed.

If the matching data exists in step S607 and the collation is OK, the recording of the attendance data,
The process of setting the collated flag, updating past data for the past N times, and updating the collation accuracy value is completed (steps S611, S612, and S613).

That is, in the fourth collation processing section 600, the fingerprint image is collated under the condition that the collation accuracy is the lowest within the allowable range. If collation still cannot be performed, the input data at that time is compared with the uncollated data. Since the data 200C is stored as the data 200C, the consistency of the collated flag of the employee master (correspondence between entering and leaving the company, etc.) is checked in a separate process for these uncollated data, and the collation target is further limited. Identify employees.

As described above, the collation operation for the input data is performed by the first collation processor 100 to the fourth collation processor 6.
00, the matching process is performed at high speed, and when the matching is performed, the date and time information is always recorded in the time and attendance data file corresponding to the matching result with the highest matching accuracy. Time recording and management with high reliability can be performed.

The third matching processing section 500 and the fourth matching processing section 600 store a single program using the matching accuracy value and the input data as parameters on a hard disk or the like. It can be loaded and run on memory as an independent process, and can also reside on memory as a single recursive and reentrant routine. With such a method, it is easy to add a lower-order collation processing unit (fifth collation processing unit).

[0115]

As described above, according to the present invention,
The time required for inputting the fingerprint image of the subject's employee can be minimized, and highly reliable working time recording can be performed.

Further, it is possible to provide a recording medium storing a program for causing a computer to perform the above-described processing.

[Brief description of the drawings]

FIG. 1 is a block diagram schematically illustrating a time recording device to which a first embodiment of the present invention has been applied.

FIG. 2 is a diagram showing a first example of the time recording device in FIG.
9 is a flowchart illustrating the operation of the matching processing unit.

FIG. 3 shows a second example of the time recording device in FIG.
9 is a flowchart illustrating the operation of the matching processing unit.

FIG. 4 is a block diagram schematically illustrating a time recording device to which a second embodiment of the present invention is applied.

FIG. 5 shows a first example of the time recording device of FIG.
9 is a flowchart illustrating the operation of the matching processing unit.

FIG. 6 shows a second example of the time recording device in FIG.
9 is a flowchart illustrating the operation of the matching processing unit.

FIG. 7 shows a third example of the time recording device of FIG.
9 is a flowchart illustrating the operation of the matching processing unit.

FIG. 8 shows a fourth example of the time recording device of FIG.
9 is a flowchart illustrating the operation of the matching processing unit.

[Explanation of symbols]

 2 Data input unit 2a Imaging unit 2b Operation unit 2C Clock unit 10, 100 First collation processing unit 20, 200 External storage device 20a, 200a · Employee master file 20b, 200b ··· Attendance data file 20C ··· Unchecked data file 200C ··· Queue data file 30 ··· Display unit 40 and 400 ··· Second collation processing unit 500 ··· 3 matching processing unit 600... 4th matching processing unit

Claims (11)

[Claims] 1. A reading means for reading a fingerprint of a subject to obtain a fingerprint image, a setting means for setting time management items,
Clocking means for measuring the date and time when the fingerprint was read,
Information input means for generating input information from a fingerprint image, management item information, and fingerprint reading date and time information obtained from each of these means; personal registration information having a registered fingerprint image stored in advance in a storage means; Matching means for identifying the subject by comparing the information with the input information from the user; and recording the management item information and date and time information input from the information input means in the time management data file of the subject identified by the matching means. A time recording device having a recording unit, wherein the matching unit has a main matching unit and a sub-matching unit, and the main matching unit stores personal registration information corresponding to input information from the information input unit. If a matching error occurs due to failure in detection, the matching operation for the input information is abandoned, while the sub-matching unit abandons the matching operation by the main matching unit. The input information, again, time recording apparatus, characterized in that for matching. 2. The main collation unit performs a collation operation for a predetermined time from a predetermined time when the frequency of information input by the information input unit increases, while the sub collation unit operates the main collation unit. 2. The time recording apparatus according to claim 1, wherein a matching operation is performed when the time recording is not performed. 3. The personal matching information is compared with the personal registration information based on code information obtained by performing predetermined image processing on a fingerprint image of the input information and encoding the personal registration information. When the collation by the means is performed, the collated personal registration information, the fingerprint image input from the information input means is recorded as a reference image, the sub collation means, the reference image and the input fingerprint image, 2. The time recording apparatus according to claim 1, wherein a comparison operation is performed by comparing. 4. The main collation unit stores the input information that could not be collated as uncollated information in the storage unit and executes a collation operation for the next input information, while the sub collation unit performs the collation operation on the next input information. Perform collation operation on collation information,
2. The time recording apparatus according to claim 1, wherein the non-collated information that has been collated is deleted from the storage unit. 5. The time recording apparatus according to claim 1, wherein said sub-collating means repeatedly executes a collating operation a plurality of times when a collation error occurs. 6. The time recording apparatus according to claim 5, wherein said sub-collating means executes a collating operation by gradually lowering the collating accuracy. 7. A notifying means for notifying the end of the collation operation, wherein said main collation means informs that the collation operation is completed irrespective of the collation result corresponding to the input information from said information input means. The time recording device according to claim 1, wherein the means is operated to notify the subject. 8. A reading means for reading a fingerprint of a subject to obtain a fingerprint image, a setting means for setting time management items,
Clocking means for measuring the date and time when the fingerprint was read,
Information input means for generating input information from a fingerprint image, management item information, and fingerprint reading date and time information obtained from each of these means; personal registration information having a registered fingerprint image stored in advance in storage means; Collating means for collating with the input information from the user to identify the subject, and recording the management item information and the date and time information inputted from the information inputting means in the time management data file of the subject identified by the collating means. A time recording device having a recording unit, wherein the matching unit is configured by a plurality of processing units having different hierarchical levels of matching accuracy and having a hierarchical structure connected in descending order of the matching accuracy. If a matching error occurs because personal registration information corresponding to the input information cannot be detected, the matching operation for the input information may be abandoned. , The input information Te transfer collation accuracy than the self matching precision to lower the lower processing means, time recording apparatus characterized by performing a matching operation for the next input information. 9. The recording means records the item information set by the setting means and the date and time information by the time keeping means in a time management data file of the subject identified by the collating means, and the personal registration information. While the collation means records the item information and the fingerprint image at the time of collation, and when collation is performed, the collation unit compares the item information recorded in the personal record information with the item information of the input information inputted this time. It is determined whether or not the personal record information has been previously collated by the lower-level processing means. If it is determined that the personal record information has been collated by the lower-level processing means, 9. The time recording apparatus according to claim 8, wherein the recording of the time management data file and the personal registration information is performed on the basis of the following. 10. The collation means, when the personal registration information corresponding to the input information is collated by the collation means, attaches an identification flag indicating that collation has been completed to the personal registration information. 9. The time recording apparatus according to claim 1, wherein a subject list that has not yet been verified is created based on the presence or absence of the flag, and the recording unit records the subject list. 11. A first step of reading a fingerprint of a subject to obtain a fingerprint image, a second step of setting time management items, a third step of measuring the date and time when the fingerprint was read, and A fourth step of identifying a subject by comparing personal registration information having a stored registered fingerprint image with the fingerprint image, management item information, and fingerprint date and time information; and time management data of the identified subject. A fifth step of recording the input management item information and date and time information in a file; and a sixth step of reducing the matching accuracy and identifying the subject again if the subject cannot be identified in the fourth step. And a computer-readable recording medium on which a program for causing a computer to execute the program is recorded.