Automatic transaction device, person detection method, and person detection program

Abstract

To provide an automatic transaction device, a person detection method, and a person detection program for detecting approach of a person to the automatic transaction device with high accuracy.SOLUTION: An automatic transaction device 10 includes: a camera 190 for photographing an image; a region detection unit 100A which detects a region of a person photographed in a photographed image by the camera 190; and an approach detection unit 100B which detects approach of a person to the automatic transaction device on the basis of a prescribed function with a factor being a value indicating the region of the person detected by the region detection unit 100A.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to an automatic transaction device, a person detection method, and a person detection program.

Background Art

[0002] There is known an automatic transaction device that transitions to a power-saving mode while there is no user performing a transaction operation. For example, Patent Document 1 describes a specific configuration of this type of automatic transaction device.

[0003] The automatic transaction device described in Patent Document 1 detects the presence or absence of a user using an optical distance sensor. This automatic transaction device transitions to the power-saving mode if no user is detected for a certain period of time, and returns to the normal operation mode when a user is detected.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, with an optical distance sensor, there is a relatively high possibility of detecting objects other than a person. Therefore, with a configuration employing an optical distance sensor, it is difficult to detect the approach of a person who is a user with high accuracy.

[0006] The present invention has been made in view of the above circumstances, and an object thereof is to provide an automatic transaction device, a person detection method, and a person detection program capable of detecting the approach of a person with high accuracy.

Means for Solving the Problems

[0007] An automatic transaction device according to an embodiment of the present invention is An imaging unit that captures an image, a range detection unit that detects the range of a person shown in the captured image captured by the imaging unit, and an approach detection unit that detects the approach of a person to the automatic transaction device. The automatic transaction device according to an embodiment of the present invention includes an imaging unit that captures an image, a range detection unit that detects the range of a person shown in the captured image captured by the imaging unit, and an approach detection unit that detects the approach of a person to the automatic transaction device. The approach detection unit determines whether the posture of the person is standing or sitting based on a first function that takes as a factor a value indicating the range of the person detected by the range detection unit. When the posture of the person is standing, a second function different from the first function, and based on the second function that takes as a factor a value indicating the range of the person detected by the range detection unit, it is detected whether the person has approached the automatic transaction device. When the posture of the person is sitting, a third function different from the first function and the second function, and based on the third function that takes as a factor a value indicating the range of the person detected by the range detection unit, it is detected whether the person has approached the automatic transaction device.

[0008] A method for detecting a person according to an embodiment of the present invention is Detect the range of a person shown in the captured image captured by the imaging unit, determine whether the posture of the person is standing or sitting based on a first function that takes as a factor a value indicating the detected range of the person. When the posture of the person is standing, a second function different from the first function, and based on the second function that takes as a factor a value indicating the detected range of the person, it is detected whether the person has approached the automatic transaction device. When the posture of the person is sitting, a third function different from the first function and the second function, and based on the third function that takes as a factor a value indicating the detected range of the person, it is detected whether the person has approached the automatic transaction device. a method for causing a computer to execute processing.

[0009] A person detection program according to an embodiment of the present invention is Detect the range of a person shown in the captured image captured by the imaging unit, determine whether the posture of the person is standing or sitting based on a first function that takes as a factor a value indicating the detected range of the person. When the posture of the person is standing, a second function different from the first function, and based on the second function that takes as a factor a value indicating the detected range of the person, it is detected whether the person has approached the automatic transaction device. When the posture of the person is sitting, a third function different from the first function and the second function, and based on the third function that takes as a factor a value indicating the detected range of the person, it is detected whether the person has approached the automatic transaction device. a program for causing a computer to execute processing.

Advantages of the Invention

[0010] According to an embodiment of the present invention, an automatic transaction device, a person detection method, and a person detection program that can detect the approach of a person with high accuracy are provided.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Embodiment for Carrying Out the Invention

[0012] Hereinafter, an automatic transaction device, a person detection method, and a person detection program according to one embodiment of the present invention will be described in detail.

[0013] FIG. 1 is a diagram showing a schematic configuration of an automatic transaction device 10 according to one embodiment of the present invention. The automatic transaction device 10 is an ATM (Automated Teller Machine) installed in a store of a financial institution such as a bank or a store of a convenience store. The automatic transaction device 10 may be referred to as, for example, a cash dispenser.

[0014] As shown in FIG. 1, the automatic transaction device 10 includes a card entrance / exit 11, a passbook entrance / exit 12, a bill entrance / exit 13, a coin entrance / exit 14, a display unit 120, and a camera 190. The card entrance / exit 11, the passbook entrance / exit 12, the bill entrance / exit 13, and the coin entrance / exit 14 are entrances / exits for inserting and discharging a cash card, a passbook, bills, and coins into and from the automatic transaction device 10, respectively.

[0015] FIG. 2 is a block diagram showing the automatic transaction device 10 according to one embodiment of the present invention. Note that the automatic transaction device 10 may have other configurations not shown in FIG. 2. That is, the form of the automatic transaction device 10 has a degree of freedom, and various design changes are possible.

[0016] The automatic transaction device 10 includes a processor 100, a memory 110, a display unit 120, an input unit 130, a passbook processing unit 140, a card processing unit 150, a bill processing unit 160, a coin processing unit 170, a communication unit 180, a camera 190, a speaker 200, and a power supply unit 210.

[0017] Processor 100 is an example of a computer that executes program 112 stored in memory 110. Processor 100 includes a RAM (Random Access Memory), a flash ROM (Read Only Memory), etc., and controls the entire automatic transaction device 10. For example, processor 100 expands various programs including program 112 stored in memory 110 onto the RAM which is a work area, and controls the automatic transaction device 10 according to the expanded programs.

[0018] Processor 100 is, for example, a single processor or a multi-processor, and includes at least one processor. When configured to include a plurality of processors, processor 100 may be packaged as a single device, or may be composed of a plurality of physically separated devices within the automatic transaction device 10.

[0019] The program 112 stored in the memory 110 causes the processor 100, which is a computer, to execute a process of detecting the range of a person captured in a captured image by a camera 190 which is an example of a photographing unit, and detecting the approach of a person to the automatic transaction device 10 based on a predetermined function using as a factor a value indicating the detected range of the person.

[0020] By executing the program 112, the approach of a person who is a user of the automatic transaction device 10 can be detected with high accuracy as compared with a conventional configuration that employs, for example, an optical distance sensor.

[0021] Processor 100 includes, as functional blocks, a range detection unit 100A, an approach detection unit 100B, a medium reception determination unit 100C, a notification unit 100D, and a mode switching unit 100E. Each functional block is realized by a program 112 executed by the processor 100 which is an example of a computer. Each functional block may be realized, in part or in whole, by hardware such as a dedicated logic circuit.

[0022] The memory 110 is a flash memory, EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), HDD (Hard Disk Drive), SSD (Solid State Drive), or the like.

[0023] The display unit 120 is, for example, an LCD (Liquid Crystal Display) equipped with a touch panel, an organic EL (Electro Luminescence), or the like. The display unit 120 displays an operation screen, a menu screen, a transaction image, etc. according to an instruction from the processor 100.

[0024] The input unit 130 is a user interface that is operated by the user and is hardware or software or a combination thereof. In this embodiment, since the display unit 120 is equipped with a touch panel, the display unit 120 also serves as the input unit 130.

[0025] The passbook processing unit 140 includes a passbook conveyance unit, a passbook sensor, a printing unit, and an MS (Magnetic Stripe) recording / playback unit. The passbook conveyance unit conveys the passbook inserted into the passbook entrance / exit 12 provided on the front of the automatic transaction device 10 into the automatic transaction device 10, and also discharges the passbook inserted into the automatic transaction device 10 from the passbook entrance / exit 12.

[0026] The passbook sensor is installed near the passbook entrance / exit 12, and detects the insertion, discharge, and receipt of the passbook by the user (in other words, the entire passbook has come out of the automatic transaction device 10 from the passbook entrance / exit 12, and for convenience, it is referred to as "receipt").

[0027] The passbook processing unit 140 notifies the processor 100 of the insertion, discharge, and receipt of the passbook detected by the passbook sensor. The printing unit prints the transaction details on the passbook conveyed into the automatic transaction device 10. The MS recording / playback unit reads and writes data to / from the magnetic stripe attached to the passbook.

[0028] The card processing unit 150 includes a card transport unit, a card sensor, a printing unit, and an MS recording / playback unit. The card transport unit transports a card (e.g., a cash card) inserted into the card entrance / exit 11 provided at the front of the automatic transaction device 10 into the automatic transaction device 10, and also discharges a card inserted into the automatic transaction device 10 from the card entrance / exit 11.

[0029] The card sensor is installed near the card entrance / exit 11, and detects the insertion, discharge, and receipt of the card by the user of the discharged card (in other words, the entire card has exited the automatic transaction device 10 from the card entrance / exit 11, and for convenience, it is referred to as "receipt").

[0030] The card processing unit 150 notifies the processor 100 of the insertion, discharge, and receipt of the card detected by the card sensor. The printing unit prints the transaction details on the card transported into the automatic transaction device 10. The MS recording / playback unit reads and writes data to / from the magnetic stripe of the card.

[0031] The bill processing unit 160 includes a temporary storage unit for temporarily storing bills, a bill discrimination unit for discriminating the authenticity, etc. of bills, a bill stacker for storing bills by type, a bill forgetting box for storing bills forgotten after the shutter of the bill entrance / exit 13 is released, a transport path for transporting bills between the bill entrance / exit 13, the temporary storage unit, the bill discrimination unit, the bill stacker, and the bill forgetting box, and the like.

[0032] The bill processing unit 160 has a bill quantity sensor for detecting the quantity of bills in the bill stacker. The bill processing unit 160 detects the quantity of bills in the bill stacker by the bill quantity sensor and notifies the processor 100. The bill quantity sensor can also detect the forgetting of bills discharged to the bill entrance / exit 13 and notify the processor 100.

[0033] The coin processing unit 170 includes a coin discrimination unit that discriminates the authenticity of coins, a coin stacker that stores coins by type, a coin forgetting box that stores coins forgotten after the shutter of the coin inlet / outlet 14 is released, a conveyance path that conveys coins between the coin inlet / outlet 14 and the coin discrimination unit, the coin stacker, and the coin forgetting box, and the like.

[0034] The coin processing unit 170 has a coin quantity sensor that detects the quantity of coins in the coin stacker. The coin processing unit 170 detects the quantity of coins in the coin stacker by the coin quantity sensor and notifies the processor 100. The coin quantity sensor can also detect the forgetting of coins discharged to the coin inlet / outlet 14 and notify the processor 100.

[0035] The communication unit 180 connects the automatic transaction device 10 to be communicable with a management system such as a host computer via a communication line of a closed network such as a dedicated line or a VPN (Virtual Private Network).

[0036] The camera 190 photographs the front area of the automatic transaction device 10 at a predetermined time interval (for example, every 200 milliseconds). The resolution of the photographed image by the camera 190 is, for example, QVGA (Quarter Video Graphics Array). The photographed image is sequentially stored in the RAM of the processor 100 in response to a command from the processor 100, and is deleted from the RAM after being used, for example, in the person detection process shown in FIG. 3 described later.

[0037] Note that the photographed image may be stored in the memory 110 at a ratio of once every n (n is a natural number) times as evidence in the case of unauthorized use or the like.

[0038] The power supply unit 210 supplies power to each block of the automatic transaction device 10. Here, the automatic transaction device 10 operates in one of a normal mode and a power saving mode.

[0039] In the normal mode, the processor 100 permits power supply to all the blocks within the automatic transaction device 10 shown in FIG. 2. That is, in the normal mode, the power supply unit 210 supplies power to all the blocks within the automatic transaction device 10 shown in FIG. 2.

[0040] In the power saving mode, the processor 100 cuts off power supply to some of the blocks within the automatic transaction device 10 shown in FIG. 2 (for example, the display unit 120, the passbook processing unit 140, the card processing unit 150, the bill processing unit 160, the coin processing unit 170, etc.) to stop the operation of those blocks. Thereby, for example, the power consumption of the automatic transaction device 10 during standby is reduced.

[0041] As described above, in the conventional configuration adopting the optical distance sensor, it is difficult to detect the approach of a person who is a user of the automatic transaction device 10 with high accuracy.

[0042] Therefore, in the automatic transaction device 10 according to the present embodiment, the person detection process described below is executed. By executing the person detection process, it is possible to avoid the problem of erroneously detecting non-persons and to detect the approach of a person who is a user of the automatic transaction device 10 with high accuracy.

[0043] FIG. 3 is a flowchart showing the person detection process by the program 112 executed by the processor 100 in an embodiment of the present invention. For example, after the power of the automatic transaction device 10 is turned on, the execution of the person detection process shown in FIG. 3 is started. The execution of the person detection process continues, for example, until the power of the automatic transaction device 10 is turned off.

[0044] As shown in FIG. 3, the program 112 acquires the captured image output from the camera 190 and stores it in the RAM of the processor 100 (step S101). Note that the captured image stored in the RAM of the processor 100 is deleted from the RAM, for example, after a predetermined time has elapsed.

[0045] Program 112 detects a person shown in the captured image (step S102). Roughly speaking, program 112 surrounds the objects in the captured image with bounding boxes, calculates the scores of the objects surrounded by the bounding boxes, and if the calculated score is equal to or greater than a predetermined threshold value, it detects this object as a person. A bounding box is a rectangular box with the vertical direction (height direction) and the horizontal direction (width direction) of the captured image as its sides. Note that for the detection process using the bounding box, for example, a method well-known in the technical field to which the invention according to the present patent application belongs is adopted. Therefore, a detailed description of the detection process in step S102 is omitted.

[0046] Even if an object surrounded by a bounding box of a small size (less than a predetermined size) is a person, the distance from the automatic transaction device 10 is far. Also, for an object surrounded by a bounding box of a small size, the accuracy of object detection and the accuracy of score calculation become low. Therefore, in step S102, for example, for an object surrounded by a bounding box of less than a predetermined size, it may not be necessary to calculate the score.

[0047] The detection process using the bounding box can also be said to be a process of detecting the range of a person shown in the captured image (exemplarily, the range indicated by the height and width of the person). Therefore, in step S102, the processor 100 that executes program 112 operates as a range detection unit 100A that detects the range of a person shown in the captured image by the camera 190, which is an example of the imaging unit.

[0048] Program 112 determines whether a person has been detected in step S102 (step S103). If a person has been detected (step S103: YES), program 112 acquires the height and width of the bounding box surrounding the detected person as the height h and width w of the person, respectively (step S104). If a person has not been detected (step S103: NO), program 112 returns to the process of step S101 and waits until the next captured image is output from the camera 190.

[0049] FIG. 4 is a diagram showing an example of a bounding box when a person is detected in a captured image. Hereinafter, for convenience, the bounding box is denoted by the symbol BB.

[0050] As shown in FIG. 4, the bounding box BB has information on height H and width W. In step S104, the height H and width W of the bounding box BB are acquired as the height h and width w of the person, respectively.

[0051] In steps S102 to S104, the processor 100 that executes the program 112 operates as a range detection unit 100A. The processor 100 operating as the range detection unit 100A detects a person captured in the captured image using the bounding box BB, and acquires the height H and width W of the bounding box BB surrounding the detected person as the height h and width w of the person, respectively. Incidentally, the range of the person is indicated by the height and width of the person captured in the captured image.

[0052] The program 112 determines whether the posture of the person detected in step S102 is a standing position based on the first function F1 (step S105).

[0053] The first function F1 is a predetermined function that uses a value indicating the range of the person detected in step S102 as a factor, and more specifically, is a linear function with the height h and width w of the person as variables. The first function F1 is represented by the following equation. f1 and C1 are the slope and intercept, respectively. The values of the slope f1 and intercept C1 are determined in advance in consideration of the statistical data of the height of Japanese people.

[0054] h = f1(w) + C1 ··· (F1)

[0055] Here, Fig. 5 shows the relationship between the height h and width w of a person shown in a photographed image and the distance between the person and the automatic transaction device 10. In Fig. 5, the vertical axis indicates the height (in other words, the number of pixels in the vertical direction of the photographed image), and the horizontal axis indicates the width (in other words, the number of pixels in the horizontal direction of the photographed image).

[0056] The 5th percentile value of the height (height from the floor to the top of the head in the standing position) of Japanese young women is 151 cm. The 95th percentile value of the height of Japanese young men is 182 cm. That is, almost all the heights of Japanese young men and women belong to the range of 151 cm to 182 cm.

[0057] In Fig. 5, the data points of black diamonds (◆) indicate the height h and width w of a person with a height of 151 cm (in other words, a person considered to have the lowest height among Japanese people) when the distance between the person and the automatic transaction device 10 is each distance D. Here, each distance D is each distance in 10 cm increments between 30 cm and 110 cm, specifically 30 cm, 40 cm, 50 cm, 60 cm, 70 cm, 80 cm, 90 cm, 100 cm, and 110 cm.

[0058] In Fig. 5, the data points of black squares (■) indicate the height h and width w of a person with a height of 182 cm (in other words, a person considered to have the highest height among Japanese people) when the distance between the person and the automatic transaction device 10 is each distance D.

[0059] Also, the 5th percentile value of the height of a Japanese young woman sitting in a wheelchair (that is, the overall sitting height indicating the height from the floor to the top of the head in the sitting position) is 120 cm. The 95th percentile value of the height of a Japanese young man sitting in a wheelchair (overall sitting height) is 138 cm. That is, almost all the heights (overall sitting height) of Japanese young men and women sitting in wheelchairs belong to the range of 120 cm to 138 cm.

[0060] In FIG. 5, the data points of the black triangles (▲) represent the height h and width w of a person with a height of 120 cm (in other words, the person who is considered to have the lowest seating overall height among Japanese people sitting in wheelchairs) and the distance between the person and the automatic transaction device 10 being each distance D (in other words, the height H and width W of the bounding box BB surrounding the person with a height of 120 cm).

[0061] In FIG. 5, the data points of the black circles (●) represent the height h and width w of a person with a height of 138 cm (in other words, the person who is considered to have the highest seating overall height among Japanese people sitting in wheelchairs) and the distance between the person and the automatic transaction device 10 being each distance D (in other words, the height H and width W of the bounding box BB surrounding the person with a height of 138 cm).

[0062] As shown in FIG. 5, the data points of the black diamonds (◆) and black squares (■) (in other words, the heights of young Japanese men and women in a standing position) are distributed in the upper region within the graph of FIG. 5. Also, the data points of the black triangles (▲) and black circles (●) (in other words, the heights of young Japanese men and women sitting in wheelchairs) are distributed in the lower region within the graph of FIG. 5.

[0063] Here, FIG. 6 visually shows the first function F1. FIG. 6 is a figure in which, with respect to FIG. 5, the first function F1 is added, the region above the first function F1 is labeled with the symbol A1, and the region below the first function F1 is labeled with the symbol A2.

[0064] As can be seen from FIG. 6, the data points of the black diamonds (◆) and black squares (■) are distributed in the upper region A1, and the data points of the black triangles (▲) and black circles (●) are distributed in the lower region A2. In other words, the data points that satisfy h≧f1(w)+C1 belong to the upper region A1, and the data points that satisfy h<f1(w)+C1 belong to the lower region A2.

[0065] Therefore, when h≧f1(w)+C1 is satisfied, Program 112 determines that the posture of the person detected in step S102 is a standing position. Also, when h<f1(w)+C1 is satisfied, Program 112 determines that the posture of the person detected in step S102 is a sitting position.

[0066] In step S105, the processor 100 that executes Program 112 operates as the approach detection unit 100B. The processor 100 operating as the approach detection unit 100B determines whether the posture of the person detected in step S102 is a standing position or a sitting position based on the first function F1.

[0067] If the posture of the person detected in step S102 is a standing position (step S105: YES), Program 112 detects whether the person has approached the automatic transaction device 10 based on the second function F2 (step S106). In the present embodiment, as an example, when the distance between the person and the automatic transaction device 10 is 50 cm or less, it is determined that the person has approached the automatic transaction device 10.

[0068] The second function F2 is a predetermined function that takes as a factor a value indicating the range of the person detected in step S102. More specifically, it is a linear function with the height h and width w of the person as variables. The second function F2 is represented by the following equation. f2 and C2 are the slope and intercept, respectively. The values of the slope f2 and the intercept C2 are also determined in advance in consideration of the statistical data of the height of Japanese people, similar to the slope f1 and the intercept C1.

[0069] h=f2(w)+C2 ···(F2)

[0070] Figure 7 visually shows the second function F2. Figure 7 is a figure in which, with respect to Figure 5, the second function F2 is added, the area above the second function F2 is labeled with symbol A3, and the area below the second function F2 is labeled with symbol A4.

[0071] As can be seen from FIG. 7, among the data points of the black diamonds (◆) and black squares (■), the data points corresponding to distances of 50 cm or less are distributed in the upper region A3, and the data points corresponding to distances exceeding 50 cm are distributed in the lower region A4. In other words, among the data points of the black diamonds (◆) and black squares (■), the data points satisfying h≧f2(w)+C2 belong to the upper region A3, and the data points satisfying h<f2(w)+C2 belong to the lower region A4.

[0072] Therefore, when h≧f2(w)+C2 is satisfied, the program 112 determines that the person detected in step S102 has approached the automatic transaction device 10. Also, when h<f2(w)+C2 is satisfied, the program 112 determines that the person detected in step S102 has not approached the automatic transaction device 10.

[0073] Thus, in step S106, the processor 100 that executes the program 112 operates as an approach detection unit 100B that detects the approach of a standing person to the automatic transaction device 10 based on the second function F2.

[0074] When the approach of a standing person to the automatic transaction device 10 is detected in step S106 (step S106: YES), the program 112 turns on a user flag indicating that there is a user at a position where the automatic transaction device 10 can be operated (step S107), returns to the process of step S101, and waits until the next captured image is output from the camera 190.

[0075] The user flag turned on here remains on as long as the approach state of the standing person to the automatic transaction device 10 is maintained (in other words, as long as the distance between the automatic transaction device 10 and the standing person is within 50 cm). This user flag is turned off, for example, when a NO determination is made in step S103 (that is, when no person is detected), or when a NO determination is made in step S106 (that is, when it is detected that the standing person in the approach state has moved away from the automatic transaction device 10).

[0076] If the posture of the person detected in step S102 is a sitting position (step S105: NO), the program 112 detects whether the person has approached the automated transaction device 10 based on the third function F3 (step S108).

[0077] The third function F3 is a predetermined function that uses as a factor a value indicating the range of the person detected in step S102. More specifically, it is a linear function with the height h and width w of the person as variables. The third function F3 is represented by the following equation. f3 and C3 are the slope and intercept, respectively. The values of the slope f3 and intercept C3 are also determined in advance in consideration of the statistical data of the height of Japanese people, similar to the slope f1 and intercept C1.

[0078] h = f3(w) + C3 ··· (F3)

[0079] Fig. 8 visually shows the third function F3. Fig. 8 is a diagram in which, compared to Fig. 5, the third function F3 is added, the area above the third function F3 is labeled with symbol A5, and the area below the third function F3 is labeled with symbol A6.

[0080] As can be seen from Fig. 8, among the data points of black triangles (▲) and black circles (●), the data points corresponding to distances of 50 cm or less are distributed in the upper region A5, and the data points corresponding to distances exceeding 50 cm are distributed in the lower region A6. In other words, among the data points of black triangles (▲) and black circles (●), the data points that satisfy h ≥ f3(w) + C3 belong to the upper region A5, and the data points that satisfy h < f3(w) + C3 belong to the lower region A6.

[0081] Therefore, when h ≥ f3(w) + C3 is satisfied, the program 112 determines that the person detected in step S102 has approached the automated transaction device 10. Also, when h < f3(w) + C3 is satisfied, the program 112 determines that the person detected in step S102 has not approached the automated transaction device 10.

[0082] Thus, in step S108, the processor 100 that executes the program 112 operates as a proximity detection unit 100B that detects the approach of a person sitting in the seat to the automatic transaction device 10 based on the third function F3.

[0083] When the approach of a person sitting in the seat to the automatic transaction device 10 is detected in step S108 (step S108: YES), the program 112 turns on the user flag (step S107), returns to the process of step S101, and waits until the next captured image is output from the camera 190.

[0084] The user flag turned on here remains on as long as the approach state of the person sitting in the seat to the automatic transaction device 10 is maintained (in other words, as long as the distance between the automatic transaction device 10 and the person sitting in the seat is within 50 cm). This user flag is turned off, for example, when the determination in step S103 is NO (that is, when no person is detected), or when the determination in step S106 is NO (that is, when it is detected that the person sitting in the seat in the approaching state has moved away from the automatic transaction device 10).

[0085] Note that when the approach of a person sitting in the seat to the automatic transaction device 10 is detected in step S108, the processor 100 may switch the screen (operation screen, menu screen, transaction image, etc.) displayed on the display unit 120 to a screen for wheelchair users. On the screen for wheelchair users, for example, operation soft buttons are laid out at the lower part (the side closer to the user) within the screen.

[0086] As described above, in the automatic transaction device 10 according to the present embodiment, by using a function with values (the height h and width w of the person indicated by the height H and width W of the bounding box BB) indicating the range of the detected person as factors, it is possible to detect with high accuracy that the person has approached the automatic transaction device 10. Further, in the automatic transaction device 10 according to the present embodiment, since an optical distance sensor is not required, the manufacturing cost can be suppressed.

[0087] In this embodiment, by utilizing the person detection result (e.g., user flag) by the program 112, it is possible to appropriately notify the user of the automatic transaction device 10 of the forgetting of media (e.g., passbook, card, banknote, coin).

[0088] FIG. 9 is a flowchart showing the media forgetting notification process by the program 112 executed by the processor 100 in an embodiment of the present invention. For example, when a predetermined transaction operation (e.g., cash withdrawal, deposit, passbook entry, balance inquiry, transfer, etc.) by the user is started, the execution of the media forgetting notification process shown in FIG. 9 is started.

[0089] As shown in FIG. 9, when the media is discharged from the automatic transaction device 10 (step S201), the program 112 determines whether the discharged media has been received based on the notifications received from various sensors (passbook processing unit 140, card processing unit 150, banknote processing unit 160, coin processing unit 170) (step S202).

[0090] In this way, in step S202, the processor 100 that executes the program 112 operates as a media reception determination unit 100C that determines whether a person who is a user of the automatic transaction device 10 has received a predetermined medium (e.g., passbook, card, banknote, coin).

[0091] Note that, in order to reduce the processing load of the processor 100 from when a predetermined transaction operation by the user is started (in other words, from when the execution of the media forgetting notification process shown in FIG. 9 is started) until the media is discharged from the automatic transaction device 10, the person detection process shown in FIG. 3 may be stopped.

[0092] When the media discharged from the automatic transaction device 10 is received (step S202: YES), the program 112 ends the media forgetting notification process shown in FIG. 9.

[0093] If the medium discharged from the automatic transaction device 10 has not been received (step S202: NO), the program 112 determines whether the user flag is in the off state (step S203). In other words, the program 112 operating as the approach detection unit 100B detects whether a person approaching the automatic transaction device 10 has moved away from the automatic transaction device 10 based on a second function (or a third function) for turning the user flag on / off.

[0094] If the user flag is in the on state (step S203: NO), it is presumed that the user who performed the transaction operation has not left in front of the automatic transaction device 10. Therefore, the program 112 returns to the process of step S202 and continues to monitor whether the medium has been received.

[0095] When the user flag becomes off (step S203: YES), it is presumed that the user who performed the transaction operation has moved away from the automatic transaction device 10 without receiving the medium. Therefore, the program 112 notifies that the user has forgotten to take the medium discharged from the automatic transaction device 10 (step S204).

[0096] In step S204, for example, a predetermined warning sound for notifying the forgetting of the medium is emitted from the speaker 200. Along with this warning sound, a predetermined warning screen for notifying the forgetting of the medium may be displayed on the display unit 120.

[0097] In this way, in step S204, when the medium reception determination unit 100C determines that the user has not received the medium and the approach detection unit 100B detects that the user has moved away from the automatic transaction device 10, the program 112 operates as a notification unit 100D for notifying the forgetting of receiving the medium.

[0098] Also, in this embodiment, by utilizing the person detection result (for example, the user flag) by the program 112, the operation mode of the automatic transaction device 10 can be appropriately switched between the normal mode and the power saving mode.

[0099] FIG. 10 is a flowchart showing mode switching processing by a program 112 executed by a processor 100 in one embodiment of the present invention. For example, after the power of the automatic transaction device 10 is turned on, the execution of the mode switching processing shown in FIG. 10 is started. The execution of the mode switching processing continues, for example, until the power of the automatic transaction device 10 is turned off. Note that at the start of the execution of the mode switching processing, the count value of a counter built in the processor 100 is reset. Also, the automatic transaction device 10 operates in the normal mode.

[0100] As shown in FIG. 10, the program 112 determines whether or not the user flag is in the on state (step S301). If the user flag is in the on state (step S301: YES), the program 112 resets the count value of the counter and continues the normal mode as the operation mode of the automatic transaction device 10 (step S302).

[0101] If the user flag is in the off state (step S301: NO), the program 112 increments the count value of the counter by 1 (step S303).

[0102] The program 112 determines whether or not the time during which no user has been in front of the automatic transaction device 10 has continued for a predetermined time (step S304). Specifically, the program 112 determines whether or not the count value of the counter has reached a predetermined value (in other words, whether or not the time during which the user flag is in the off state has continued for a predetermined time).

[0103] If the time during which no user has been in front of the automatic transaction device 10 has not continued for a predetermined time (step S304: NO), the program 112 returns to the process of step S301 and continues to monitor the user flag.

[0104] When the time during which there is no user in front of the automatic transaction device 10 continues for a predetermined time (step S304: YES), the program 112 resets the count value of the counter and switches the operation mode of the automatic transaction device 10 to the power saving mode (step S305). As long as the user flag is in the off state (step S306: NO), the program 112 continues the power saving mode.

[0105] When the user flag becomes on (step S306: YES), the program 112 resets the count value of the counter and switches the operation mode of the automatic transaction device 10 to the normal mode (step S302), returns to the process of step S301, and continues to monitor the user flag.

[0106] As described above, when performing the mode switching process shown in FIG. 10, the processor 100 that executes the program 112 operates as a mode switching unit 100E that switches the operation mode of the automatic transaction device 10 between the normal mode and the power saving mode. The processor 100 operating as the mode switching unit 100E switches the operation mode of the automatic transaction device 10 according to the detection result by the proximity detection unit 100B (exemplarily, the user flag).

[0107] Note that the present invention is not limited to the above-described embodiments as they are, and the components can be modified and embodied without departing from the gist thereof at the implementation stage. Also, various inventions can be formed by appropriately combining a plurality of components disclosed in the above embodiments. For example, all the components shown in the embodiments may be appropriately combined. Further, components from different embodiments may be appropriately combined. Various modifications and applications are possible within the scope not departing from the gist of the invention.

[0108] In the above embodiment, the presence or absence of a user of the automatic transaction device 10 is detected based on the person detection result for one captured image. However, in another embodiment, the presence or absence of a user of the automatic transaction device 10 may be detected based on the person detection results of a plurality of continuously captured images.

[0109] By comparing the sizes of the bounding boxes in each of a plurality of continuously captured images, it is possible to estimate that the person surrounded by the bounding box is gradually approaching the automatic transaction device 10, gradually moving away from the automatic transaction device 10, or stationary near the automatic transaction device 10, etc.

[0110] For example, when it is estimated that a standing person is gradually approaching the automatic transaction device 10, for example, in step S106 of FIG. 3, the program 112 may detect whether the person has approached the automatic transaction device 10 based on the fourth function F4. The fourth function F4 is a linear function whose slope and intercept are different from those of the second function F2. For example, when the distance between the standing person and the automatic transaction device 10 is 100 cm or less, it is a function for determining that the person has approached the automatic transaction device 10.

Explanation of Signs

[0111] 10: Automatic transaction device 11: Card entrance / exit 12: Passbook entrance / exit 13: Bill entrance / exit 14: Coin entrance / exit 100: Processor 100A: Range detection unit 100B: Approach detection unit 100C: Medium reception determination unit 100D: Notification unit 100E: Mode switching unit 110: Memory 112: Program 120: Display unit 130: Input unit 140: Passbook processing unit 150: Card processing unit 160: Bill processing unit 170: Coin processing unit 180: Communication unit 190: Camera 200: Speaker 210: Power supply unit

Claims

1. An imaging unit that captures an image, A range detection unit that detects the range of a person shown in the captured image captured by the imaging unit, An approach detection unit that detects the approach of the person to the automatic transaction device, and includes, The approach detection unit, Based on a first function that uses a value indicating the range of the person detected by the range detection unit as a factor, determines whether the posture of the person is standing or sitting, When the posture of the person is standing, a second function different from the first function, and based on a second function that uses a value indicating the range of the person detected by the range detection unit as a factor, detects whether the person has approached the automatic transaction device, When the posture of the person is sitting, a third function different from the first function and the second function, and based on a third function that uses a value indicating the range of the person detected by the range detection unit as a factor, detects whether the person has approached the automatic transaction device, An automatic transaction device.

2. The range of the person is indicated by the height and width of the person shown in the captured image, The automatic transaction device according to claim 1.

3. The first function, the second function, and the third function are linear functions having the height and width of the person as variables, The automatic transaction device according to claim 2.

4. The range detection unit detects the person using a bounding box, and acquires the height and width of the bounding box surrounding the detected person as the height and width of the person, respectively, The automatic transaction device according to claim 2 or claim 3.

5. A media reception determination unit that determines whether the person has received a predetermined medium, A notification unit that notifies of forgetting to receive the medium, and further includes, The approach detection unit detects whether the person who has approached the automatic transaction device has moved away from the automatic transaction device based on the second function or the third function, When the media reception determination unit determines that the person has not received the medium and the approach detection unit detects that the person has moved away from the automatic transaction device, the notification unit notifies of forgetting to receive the medium, The automatic transaction device according to any one of claims 1 to 4.

6. The automatic transaction device further includes a mode switching unit that switches the operation mode of the automatic transaction device between a normal mode and a power saving mode, The mode switching unit switches the operation mode according to the detection result by the approach detection unit. The automatic transaction device according to any one of claims 1 to 5.

7. Detect the range of a person shown in a captured image by a photographing unit, Based on a first function using as a factor a value indicating the range of the detected person, determine whether the posture of the person is standing or sitting, When the posture of the person is standing, based on a second function different from the first function and using as a factor a value indicating the range of the detected person, detect whether the person has approached the automatic transaction device, When the posture of the person is sitting, based on a third function different from the first function and the second function and using as a factor a value indicating the range of the detected person, detect whether the person has approached the automatic transaction device, and cause a computer to execute the process. Person detection method.

8. Detect the range of a person shown in a captured image by a photographing unit, Based on a first function using as a factor a value indicating the range of the detected person, Determine whether the posture of the person is standing or sitting, When the posture of the person is standing, based on a second function different from the first function and using as a factor a value indicating the range of the detected person, detect whether the person has approached the automatic transaction device, When the posture of the person is sitting, based on a third function different from the first function and the second function and using as a factor a value indicating the range of the detected person, detect whether the person has approached the automatic transaction device, and cause a computer to execute the process. Person detection program.

Images