Greeting system

Abstract

To greet a visitor with multiple voices, to produce a lively atmosphere.SOLUTION: A greeting system 1 which greets a customer C who visits a store S includes: a camera 11 which images the customer at an entrance Dw of the store; a memory 32 of a computer 30 that stores a plurality of first sound sources related to a first greeting "Hello, may I help you?" for the customer C; a CPU 31 of the computer 30 that reproduces the stored first sound sources when a detection unit detects the customer; and speakers 21, 22 which are installed in a staffed area E2 in the store S and output voices based on the reproduced first sound sources.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a greeting providing system that provides greetings to customers who visit a store. [Background technology]

[0002] Conventionally, in stores such as restaurants, greetings such as "Welcome" and "Thank you" are given to customers. Patent documents disclose devices that improve crime prevention in stores by automatically generating such greeting sounds. The voice generating device according to the patent documents includes a sensor that detects a human body, a speaker that outputs voice, and a main body in which a plurality of phrases are recorded in advance, an interval time is selected for each phrase, and the main body controls the interval time between the detection of the human body and the utterance of the recorded phrase. When the voice generating device detects a human body with the sensor, it utters a first phrase from the speaker, and utters the next phrase after a predetermined interval time has elapsed since the utterance of the phrase ended. Here, "Welcome," "How is it?", and "How is it?" are given as examples of the plurality of phrases. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2002-182696 A Summary of the Invention [Problem to be solved by the invention]

[0004] The voice generating device in the above-mentioned patent document produces a plurality of pre-recorded phrases at a preset timing. Therefore, customers staying in the store hear the same phrases repeatedly every time a new customer enters the store, which makes them feel that the phrases are mechanically outputted impersonal greetings, making it difficult to create a lively atmosphere in the store. [Means for solving the problem]

[0005] In order to solve the above problems, the present invention is a greeting provision system that provides a greeting to a customer who visits a store, and is characterized by comprising: a detection unit that detects a customer at a first location; a memory unit that stores a plurality of first sound sources related to a first greeting to the customer; a playback unit that plays the stored plurality of first sound sources when the customer is detected by the detection unit; and an audio output unit that is installed in an area in the store where staff are located and outputs audio based on the played plurality of first sound sources.

[0006] In the present invention, the reproduction unit changes the reproduction mode of the plurality of first sound sources every time the detection unit detects a customer.

[0007] In the present invention, the audio output unit has a plurality of speakers, and the reproduction unit inputs audio signals generated based on different first sound sources to each of the speakers.

[0008] In addition, in the present invention, the detection unit detects a customer at a second location different from the first location, the memory unit stores a plurality of second sound sources related to a second greeting for the customer at the second location, the playback unit plays the stored plurality of second sound sources when a customer at the second location is detected by the detection unit, and the audio output unit outputs audio based on the played plurality of second sound sources. Effect of the Invention

[0009] According to the present invention, a lively atmosphere can be created inside the store. [Brief description of the drawings]

[0010] [Figure 1] A floor plan of a restaurant in which a greeting providing system according to an embodiment of the present invention is installed. [Diagram 2] Side view of the restaurant [Diagram 3] Hardware configuration of the greeting provision system [Figure 4] Flow diagram of the greeting provision system DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, a greeting providing system according to an embodiment of the present invention will be described using a restaurant as an example.

[0012] As shown in FIG. 1 to FIG. 3, the greeting providing system 1 according to the present embodiment is installed in a restaurant S (hereinafter, also referred to as "store S"). The restaurant S is provided with an eating area E1 that functions as a customer area where products and services are provided to customers C who visit the restaurant, and a cooking area E2 that functions as a staff area used only by staff and where customers C are basically not allowed to enter (use). The eating area E1 is an area where food and drink are provided to customers C, and the customers C eat the provided food and drink. The eating area E1 is provided with a plurality of tables and a plurality of chairs corresponding to each table, and is provided with an entrance / exit Dw through which customers C enter and exit the restaurant, and a return port Rs through which tableware is returned. The cooking area E2 is an area where staff cook the food and drink to be provided to customers C, and is provided with various cooking equipment.

[0013] The greeting providing system 1 installed in the restaurant S is a system that provides greetings and calls (collectively referred to as "greetings") to customers C who visit the restaurant, and is equipped with two cameras 11, 12, two speakers 21, 22, and a computer 30.

[0014] One of the cameras 11 (hereinafter referred to as the "first camera 11") functions as a first imaging unit that images a customer C at an entrance / exit Dw (first location) of the store S. The first camera 11 is installed on the ceiling in the store S and is arranged so as to include the entrance / exit Dw of the store S in its angle of view. The other of the cameras 12 (hereinafter referred to as the "second camera 12") functions as a second imaging unit that images a customer C at a tableware return slot Rs (second location). The second camera 12 is installed on the ceiling in the store S and is arranged so as to include the return slot Rs in its angle of view. The first camera 11 and the second camera 12 capture images of target locations at a predetermined frame rate to generate images. The first camera 11 and the second camera 12 are connected to the computer 30 so as to be able to communicate with each other by wire or wirelessly, and input the generated images to the computer 30.

[0015] Speakers 21, 22 (hereinafter also referred to as "first speaker 21" and "second speaker 22") function as an audio output unit that outputs audio related to greetings (hereinafter referred to as "greeting audio"). Speakers 21, 22 are installed on the ceiling of cooking area E2 and positioned to output audio toward eating area E1. Each of the speakers is connected to computer 30 by wire and outputs audio based on an audio signal output from computer 30.

[0016] The computer 30 includes a CPU 31 and a memory 32. A greeting providing program including the processing described below is stored in advance in the memory 32, and the CPU 31 executes the greeting providing program. As a result, the computer 30 functions as a recognition unit that recognizes the customer C at a predetermined location (the entrance / exit Dw and the return exit Rs) in the store S based on the images input from the first camera 11 and the second camera 12, and as a playback unit that plays a greeting voice based on the recognition result.

[0017] Here, the memory 32 of the computer 30 stores a group of sound sources corresponding to two types of greeting voices for the customer C. Specifically, a first group of sound sources corresponding to a greeting when the customer C enters the store (hereinafter referred to as a "first greeting"), and a second group of sound sources corresponding to a greeting expressing gratitude to the customer C (hereinafter referred to as a "second greeting") are stored.

[0018] The first sound source group includes a first sound source A, a first sound source B, and a first sound source C. The first sound source A is a sound source generated by recording a male voice A of the first greeting "Welcome" spoken by a male A. The first sound source B is a sound source generated by recording a male voice B of the first greeting "Welcome" spoken by a different male B. The first sound source C is a sound source generated by recording a female voice C of the first greeting "Welcome, hello" spoken by a female C.

[0019] The second sound source group includes second sound source A, second sound source B, and second sound source C. Second sound source A is a sound source generated by recording male voice A of the second greeting "Thank you very much" spoken by male A. Second sound source B is a sound source generated by recording male voice B of the second greeting "Thank you very much" spoken by male B. Second sound source C is a sound source generated by recording female voice C of the second greeting "Thank you very much, please come again" spoken by female C.

[0020] In this way, the memory 32 of the computer 30 functions as a storage unit that stores a sound source corresponding to the greeting voice.

[0021] The greeting providing system 1 configured as above executes a detection process s10 and a reproduction process s20 in this order, as shown in FIG.

[0022] The detection process s10 is a process for detecting the customer C at each location (the entrance / exit Dw and the return exit Rs) in the store S, and includes an imaging process s11 and a recognition process s12.

[0023] The imaging process s11 includes a first imaging process and a second imaging process. The first imaging process is a process in which the first camera 11 captures an image of the entrance / exit Dw of the store S and generates an image of the image. The image generated in this process is input to the computer 30 each time. The second imaging process is a process in which the second camera 12 captures an image of the return exit Rs and generates an image of the image. The image generated in this process is input to the computer 30 each time. The imaging process s11 is executed at a predetermined frame rate.

[0024] The recognition process s12 is a process in which the CPU 31 of the computer 30 recognizes the customer C at each location based on each image input from the first camera 11 and the second camera 12 (hereinafter, referred to as the "first input image" and the "second input image", respectively, and collectively referred to as the "input images"). The process for recognizing the customer C is not particularly limited, but in this embodiment, when a person is recognized by analyzing each input image, the person is recognized as the customer C. Specifically, the recognition process s12 includes a difference process and a human body recognition process. In the difference process, an image when no person is present at the entrance / exit Dw (hereinafter referred to as a "first unmanned image") and an image when no person is present at the return exit Rs (hereinafter referred to as a "second unmanned image") are stored in advance in the memory 32. Then, the CPU 31 takes the difference between the first input image and the first unmanned image to generate a first difference image. As a result, when a customer C is present at the entrance / exit Dw, the image of the customer C is extracted as the first difference image. Furthermore, the CPU 31 takes the difference between the second input image and the second unmanned image to generate a second difference image. As a result, when a customer C is present at the return exit Rs, the image of the customer C is extracted as the second difference image. The human body recognition process is a process for recognizing the presence or absence of a human body based on each difference image. For example, the process recognizes the presence or absence of a human body using machine learning. Specifically, machine learning is performed using images related to the human body as training data to generate a trained model in advance, the difference images are input to the trained model, and the presence or absence of a human body is determined based on the output of the trained model. Alternatively, the presence or absence of a human body may be recognized by defining feature points related to the human body in advance, comparing the feature points in the difference image with the feature points related to the human body, and determining whether the feature points are similar to each other. In this way, the first camera 11 functioning as the first imaging unit and the computer 30 functioning as the recognition unit function as a first detection unit that detects the customer C at the entrance / exit Dw of the store S. In addition, the second camera 12 functioning as the second imaging unit and the computer 30 functioning as the recognition unit function as a second detection unit that detects the customer C at the return exit Rs of the store S.

[0025] The playback process s20 is a process for playing back the sound sources stored in the memory 32 based on the result of the detection process s10. Specifically, the CPU 31 of the computer 30 checks the result of the detection process s10 (hereinafter referred to as the "detection result"). If the customer C is not detected in the detection result, the process returns to the imaging process s11. On the other hand, if the customer C is detected in the detection result, the sound sources stored in the memory 32 are played back, and audio signals corresponding to the sound sources are output to the speakers 21 and 22.

[0026] For example, when a customer C is detected at the entrance / exit Dw of the store S, the first sound source group is played. Here, the computer 30 outputs audio signals corresponding to the first sound source A and the first sound source C to the first speaker 21, and outputs an audio signal corresponding to the first sound source B to the second speaker 22. As a result, a male voice A and a female voice C corresponding to the first greeting are generated from the first speaker 21, and a male voice B corresponding to the first greeting is generated from the second speaker 22.

[0027] Furthermore, when customer C is detected at the return slot Rs, the second sound source group is played. Then, computer 30 outputs audio signals corresponding to second sound source A and second sound source B to first speaker 21, and outputs an audio signal corresponding to second sound source C to second speaker 22. As a result, male voice A and male voice B corresponding to the second greeting are generated from first speaker 21, and female voice C corresponding to the second greeting is generated from second speaker 22.

[0028] According to the greeting providing system 1 of the present embodiment, when the customer C enters the store S, a first greeting can be issued to the customer C, and when the customer C returns the tableware, a second greeting can be issued to the customer C. Here, each greeting is output by a plurality of voices, so that it is possible to create a lively atmosphere in the store S. Furthermore, since the speakers 21, 22 are installed in the cooking area E2, it is possible to make the customer feel as if the greeting is being delivered by the cooking staff.

[0029] Although the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment and may be modified as follows.

[0030] <Variation 1> The greeting providing system 1 of the above embodiment includes two speakers 21, 22 installed in the cooking area E2, but is not limited to this aspect. That is, the greeting providing system 1 may include one speaker installed in the cooking area E2. Also, the greeting providing system 1 may include three or more speakers installed in the cooking area E2. In a modified aspect including multiple speakers, it is preferable that the computer 30 outputs audio signals so that different sounds are output from each of the speakers. For example, in an aspect including three speakers, the audio signal of the first sound source A is output to the first speaker, the audio signal of the first sound source B is output to the second speaker, and the audio signal of the first sound source C is output to the third speaker.

[0031] <Variation 2> In the above embodiment, the first sound source and the second sound source are sound sources generated by recording a single voice such as a male voice or a female voice, but they may be sound sources generated by recording the voices of multiple people (composite voices). Specifically, the first sound source A may be a sound source generated by recording a composite voice A of the first greeting pronounced by multiple people. The first sound source B may be a sound source generated by recording a composite voice B of the first greeting pronounced by multiple people different from the first sound source A. Similarly, the first sound source C is a sound source generated by recording a composite voice C of the first greeting pronounced by multiple people different from the first sound source A and the first sound source B.

[0032] <Variation 3> In the above embodiment, three first sound sources are stored in the first sound source group and three second sound sources are stored in the second sound source group in the memory 32, but the first sound source group and the second sound source group may store more sound sources.

[0033] <Variation 4> In the reproduction process s20 of the above embodiment, the CPU 31 of the computer 30 may change the reproduction modes of a plurality of sound sources.

[0034] <Modification 4-1> For example, the CPU 31 of the computer 30 may randomly play back multiple sound sources. Specifically, every time a customer C is detected at the entrance / exit Dw, the CPU 31 of the computer 30 outputs a sound signal of a first sound source randomly selected from the first sound source group in the memory 32 to each speaker. Similarly, when a customer C is detected at the return exit Rs, the CPU 31 of the computer 30 outputs a sound signal of a second sound source randomly selected from the second sound source group in the memory 32 to each speaker. In this way, by randomly playing back multiple sound sources, a different sound is output to the customer C each time. Therefore, a natural greeting can be provided.

[0035] <Modification 4-2> Also, for example, in the playback process s20 of the above embodiment, multiple sound sources may be played back at different timings. Specifically, every time a customer C is detected at the entrance Dw, the CPU 31 of the computer 30 plays back multiple first sound sources selected from the first sound source group in the memory 32 at different timings. Furthermore, the CPU 31 of the computer 30 may set the playback timings randomly. In this way, by outputting the sound at different timings, a natural greeting can be provided.

[0036] <Variation 5> In the recognition process s12 of the above embodiment, the discrimination process may be performed after the human body recognition process. The discrimination process is a process for discriminating between the customer C and the staff. Specifically, machine learning is performed using an image of the staff as training data to generate a trained model, and the difference image is input to the trained model. If the output of the trained model indicates a staff member, the CPU 31 of the computer 30 does not execute the reproduction process s20. On the other hand, if the output of the trained model does not indicate a staff member, the CPU 31 of the computer 30 executes the reproduction process s20. The discrimination process may involve determining whether or not feature points related to staff members are included in a differential image recognized as a person by predetermining feature points related to the staff members. If the result of the determination is that the feature points related to the staff members are not included, the image is determined to be customer C. On the other hand, if the feature points related to the staff members are included, the image is determined to be staff. Only when the result of the discrimination process is that the image is determined to be customer C, the CPU 31 of the computer 30 plays the sound source in the playback process s20.

[0037] <Variation 6> In the recognition process s12 of the above embodiment, a re-entry detection process may be executed. The re-entry detection process is a process for detecting the re-entry of a customer C within a certain period of time, and is executed after the human body recognition process or the distinction process. The re-entry detection process executes a storage process for storing a difference image obtained by the difference process in the memory 32 for a certain period of time, and calculates the degree of coincidence of feature points by comparing the difference image with other difference images stored in the memory 32. If the degree of coincidence is high, it is determined that the same customer C has re-entered the store within the certain period of time. If it is determined that the customer C has re-entered the store in this way, the greeting has already been provided, so the CPU 31 of the computer 30 does not execute the reproduction process s20.

[0038] <Variation 7> In the above embodiment, the first camera 11 and the computer 30 function as the first detector, and the second camera 12 and the computer 30 function as the second detector, but the present invention is not limited to this embodiment. For example, a single camera may capture the entrance / exit Dw and the return exit Rs to generate an image, and the computer 30 may detect a customer at the entrance / exit Dw and the return exit Rs based on the image. In other words, a single camera and the computer 30 may function as the detector.

[0039] <Variation 8> In the above embodiment and modification 7, the camera and computer 30 function as the detection unit, but the detection unit may be a human sensor. The human sensor is a sensor that operates in response to a human body at a predetermined location (entrance / exit Dw and return exit Rs), and may be, for example, an infrared sensor that detects a human body at a predetermined location by infrared rays emitted from the human body, a sound sensor that detects a human body at a predetermined location by changes in sound, or an ultrasonic sensor that transmits ultrasonic waves from a transmitter to a predetermined location and receives the ultrasonic waves reflected from the human body with a receiver to detect a human body at a predetermined location. The output of the human sensor is electrically connected to the computer 30, and a detection signal when a human body is detected is input to the computer 30.

[0040] In such a modified embodiment, the detection process s10 in the above process flow is executed in the human sensor. In the playback process s20, the CPU 31 of the computer 30 checks whether or not there is a detection signal from the human sensor. Then, when the detection signal is confirmed, the CPU 31 of the computer 30 plays back the multiple sound sources stored in the memory 32.

[0041] In addition, if the door of the store S is an automatic door, a sensor that detects a human body to open and close the door may be used as the detection unit.

[0042] <Variation 9> In the above embodiment, the customer C is detected at the entrance Dw and the return exit Rs of the store S, but it is also possible to detect only one of them.

[0043] <Modification 10> Although the greeting providing system 1 of the above embodiment is installed in a restaurant S, it can also be used in stores S other than restaurants S, such as shops, salons, supermarkets, convenience stores, shopping malls, and the like. [Explanation of symbols]

[0044] 1. Greetings Providing System 11 Camera 1 12 Second Camera 21 First Speaker 22 Second Speaker 30 Computers

Claims

1. A greeting providing system for providing greetings to customers who visit a store, a detector for detecting a customer at a first location; A storage unit that stores a plurality of first sound sources related to a first greeting to the customer; a playback unit that plays the stored first sound sources when the customer is detected by the detection unit; a sound output unit that is installed in an area where staff are located in the store and outputs sounds based on the reproduced first sound sources; A greeting providing system comprising:

2. The greeting providing system according to claim 1 , wherein the playback unit changes a playback mode of the plurality of first sound sources every time the detection unit detects a customer.

3. The audio output unit has a plurality of speakers, The greeting providing system according to claim 1 , wherein the reproduction unit inputs, to each of the speakers, a sound signal generated based on a different first sound source.

4. The detection unit detects a customer at a second location different from the first location, The storage unit stores a plurality of second sound sources related to a second greeting to the customer at the second location; The playback unit plays the stored second sound sources when the detection unit detects a customer at the second location, and The greeting providing system according to claim 1 , wherein the audio output unit outputs audio based on the reproduced second sound sources.

Images