Information processing system, information processing method, program, and recording medium

Abstract

The present invention comprises: an image acquisition unit that acquires an image captured by an imaging unit which moves together with a moving body; a criteria setting unit that sets criteria for images to be stored, on the basis of a language input by a user; a matching degree calculation unit that calculates, for one image acquired by the image acquisition unit, a matching degree relative to the criteria; a determination unit that determines whether or not to store the image on the basis of the matching degree pertaining to the image; and a storage control unit that stores, in a storage unit, an image which is determined by the determination unit to be stored.

Description

Information processing system, information processing method, program, and recording medium

[0001] The present invention relates to an information processing system, an information processing method, a program, and a recording medium.

[0002] An in-vehicle device that captures a subject that matches the preferences of a vehicle occupant has been disclosed. For example, Patent Literature 1 discloses an image capturing device that collects images uploaded to a social networking site by the vehicle occupant, estimates a subject that the occupant likes from the collected images, and, if a natural object or an artificial object that matches the estimated subject is recognized while capturing a scene outside the vehicle, displays an image that includes the natural object or the artificial object on a display.

[0003] Japanese Patent Application Laid-Open No. 2019-114875

[0004] In the image capturing device disclosed in Patent Document 1, the passenger's favorite subject is automatically estimated only from images previously uploaded to a social networking site. Therefore, there is a problem in that the passenger cannot freely specify the subject, for example, a subject that has not been photographed before will not be photographed even if the passenger is interested in it.

[0005] The present invention aims to solve the above-mentioned problems and to provide an information processing system, an information processing method, a program, and a recording medium that can automatically store images of subjects or scenes that a user likes while a moving object is moving.

[0006] The information processing system described in claim 1 is characterized by having an image acquisition unit that acquires images captured by an imaging unit that moves with the moving object, a standard setting unit that sets standards for images to be stored based on language input by the user, a compatibility calculation unit that calculates the compatibility of one image acquired by the image acquisition unit with the standards, a judgment unit that determines whether or not one image should be stored based on the compatibility of the one image, and a memory control unit that stores images that the judgment unit determines should be stored in the memory unit.

[0007] The information processing method described in claim 10 is an information processing method executed by an information processing system, and is characterized by having an image acquisition step of acquiring an image captured by an imaging unit moving together with the moving body, a standard setting step of setting standards for images to be stored based on language input by the user, a compatibility calculation step of calculating the compatibility of one image acquired in the image acquisition step with the standards, a judgment step of determining whether or not one image should be stored based on the compatibility of the one image, and a memory control step of storing in a memory unit the image determined to be stored in the judgment step.

[0008] The program described in claim 11 is a program to be executed by a computer, and is characterized in that it causes the computer to execute the following steps: an image acquisition step of acquiring an image captured by an imaging unit that moves together with the moving body; a standard setting step of setting standards for images to be stored based on language input by the user; a compatibility calculation step of calculating the compatibility of one image acquired in the image acquisition step with the standards; a determination step of determining whether or not one image should be stored based on the compatibility of the one image; and a storage control step of storing in a storage unit an image determined to be stored in the determination step.

[0009] The recording medium described in claim 12 is a recording medium having recorded thereon a program that causes a computer provided in an information processing system to execute the following steps: an image acquisition step of acquiring an image captured by an imaging unit that moves together with the moving object; a standard setting step of setting standards for images to be stored based on language input by a user; a suitability calculation step of calculating the suitability of one image acquired in the image acquisition step with respect to the standards; a determination step of determining whether or not one image should be stored based on the suitability of the one image; and a storage control step of storing in the storage unit an image determined in the determination step to be stored.

[0010] 1 is a diagram illustrating a configuration of an automatic photography system according to a first embodiment; FIG. 2 is a diagram illustrating a configuration of a front seat portion of a vehicle and a state inside the vehicle according to a first embodiment; FIG. 3 is a block diagram illustrating an example of a configuration of an in-vehicle device according to a first embodiment; FIG. 4 is a block diagram illustrating an example of a configuration of a server device according to a first embodiment; FIG. 5 is a diagram illustrating an example of caption generation of an image performed by the server device according to a first embodiment; FIG. 6 is a diagram illustrating an example of calculation of an aesthetic evaluation value of an image performed by the server device according to a first embodiment; FIG. 7 is a diagram illustrating an example of a flowchart of a control routine executed by the in-vehicle device according to a first embodiment; FIG. 8 is a diagram illustrating an example of a flowchart of a control routine executed by the in-vehicle device according to a first embodiment; 10A and 10B are diagrams illustrating an example of a waveform of an aesthetic evaluation value in the automatic photography system according to the third embodiment and a waveform obtained by smoothing the waveform.

[0011] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same components are designated by the same reference numerals, and the description of the same components will be omitted.

[0012] First, the overall system configuration will be described. Fig. 1 is a diagram showing the configuration of an automatic photography system 100 according to Example 1. As shown in Fig. 1, the automatic photography system 100 includes an in-vehicle device 10 mounted on a vehicle M as a moving body, and a server device 11.

[0013] In this embodiment, the in-vehicle device 10 and the server device 11 can transmit and receive data to and from each other via a network NW using a communication protocol such as TCP / IP or UDP / IP. The connection between the in-vehicle device 10 and the network NW can be established by mobile communication such as 3G (3rd Generation), 4G (4th Generation), or 5G (5th Generation), or wireless communication such as Wi-Fi (registered trademark).

[0014] 2 is a diagram showing the configuration of the front seat area of ​​the vehicle M and the interior of the vehicle M. The in-vehicle device 10 is connected to an exterior camera 13, a microphone 14, a touch panel display 15, and a speaker 16 installed in the vehicle M, and includes a control unit that controls these. The in-vehicle device 10 is disposed, for example, in the center of a dashboard DB in the front seat of the vehicle M.

[0015] The exterior camera 13 is an imaging device that captures images of the situation outside the vehicle M. The exterior camera 13 is a wide-angle camera that captures images of the area in front of the vehicle M through the front windshield FW. The exterior camera 13 is installed, for example, on the ceiling inside the vehicle M, adjacent to the rearview mirror RM.

[0016] The microphone 14 is a voice input device that receives sounds inside the vehicle, such as the voice of the driver DR of the vehicle M, and converts them into an electrical signal. The microphone 14 is disposed, for example, on a dashboard DB in the front seat of the vehicle M.

[0017] The touch panel display 15 is a display device that combines a display that displays a screen based on the control of the in-vehicle device 10 and a touch panel that accepts input operations from the passengers of the vehicle M. The touch panel display 15 displays, for example, a navigation image in which the current position of the vehicle M is superimposed on a map. In this embodiment, the touch panel display 15 is disposed in the center of the dashboard DB.

[0018] The speaker 16 is an audio output device that outputs audio into the vehicle cabin based on the control of the in-vehicle device 10. The speaker 16 is provided, for example, on each of the two A-pillars AP of the vehicle M.

[0019] Note that the above-described locations of the in-vehicle device 10, the exterior camera 13, the microphone 14, the touch panel display 15, and the speaker 16 in the front seat area of ​​the vehicle M are merely examples, and they may be arranged in other locations.

[0020] In the following description, it is assumed that the vehicle M is scheduled to pass through a natural area, such as mountains and a lake, while traveling from a departure point of the vehicle M to a destination. Fig. 1 shows an example of a case where a user makes a statement to the in-vehicle device 10. For example, a passenger of the vehicle M (e.g., the driver DR) wants to keep a landscape photo that includes both mountains and a lake while traveling in the vehicle M, and utters the words "Take a photo that includes both mountains and a lake!" into the microphone 14 to specify the conditions for the image to be stored. Note that the condition may be specified verbally before the vehicle M starts traveling.

[0021] The in-vehicle device 10 is configured to be capable of speech recognition of speech (natural language) from a user of the in-vehicle device 10, for example, a driver DR of the vehicle M. That is, the in-vehicle device 10 converts the speech input by the driver DR of the vehicle M into text consisting of character strings.

[0022] In the automatic photography system 100, the in-vehicle device 10 converts the voice input by the driver DR of the vehicle M into text, then extracts words of a specified part of speech from the text, and uses the extracted words to set criteria as conditions for the scene of the image to be stored.

[0023] In addition, in the automatic photography system 100, the server device 11 acquires images captured by the exterior camera 13 of the vehicle-mounted device 10 and generates a sentence, a so-called caption, indicating the scene depicted in the acquired image.

[0024] In the automatic photography system 100, the server device 11 acquires images captured by the exterior camera 13 of the in-vehicle device 10, and calculates an aesthetic evaluation value that indicates the beauty of the acquired images.

[0025] In the automatic photography system 100, the vehicle-mounted device 10 calculates the degree of conformance between a caption generated by the server device 11 for one image taken by the exterior camera 13 and the criteria for images to be stored set by the vehicle-mounted device 10, and if it determines that the calculated degree of conformance is above a predetermined level and that the aesthetic evaluation value calculated by the server device 11 for one image is above a predetermined level, it stores the one image.

[0026] [Configuration of In-Vehicle Device 10] Next, the configurations of the in-vehicle device 10 and the server device 11 will be described. First, the configuration of the in-vehicle device 10 will be described with reference to Fig. 3. Fig. 3 is a block diagram showing an example of the configuration of the in-vehicle device 10 according to the first embodiment. The in-vehicle device 10 is a device in which a control unit 18, a communication unit 19, an input unit 21, an output unit 22, and a large-capacity storage device 23 cooperate with each other via a system bus 17.

[0027] The control unit 18 is configured with a CPU (Central Processing Unit) 18A, a ROM (Read Only Memory) 18B, a RAM (Random Access Memory) 18C, etc., and functions as a computer. The CPU 18A reads and executes various programs stored in the ROM 18B and the large-capacity storage device 23, thereby realizing various functions.

[0028] The various programs may be acquired, for example, from another server device or the like via a network, or may be recorded on a recording medium and read via a drive device. That is, the various programs stored in the mass storage device 23 (including a program for executing processing in the in-vehicle device 10, which will be described later) can be transmitted via a network, or can be recorded on a computer-readable recording medium and transferred.

[0029] The communication unit 19 is a communication device that transmits and receives data to and from external devices in accordance with instructions from the control unit 18. The communication unit 19 is, for example, a network interface card (NIC) for connecting to a network.

[0030] For example, the control unit 18 acquires images taken by the outside-vehicle camera 13 at certain time intervals and transmits the acquired images to the server device 11 via the communication unit 19. In other words, the control unit 18 functions as an image acquisition unit that acquires images taken by the outside-vehicle camera 13 as an imaging unit.

[0031] The input unit 21 is an interface unit that is communicatively connected between the control unit 18 of the in-vehicle device 10 and the exterior camera 13, microphone 14, and touch panel display 15. In the automatic photography system 100, the control unit 18 acquires the voice of the driver DR of the vehicle M input from the microphone 14 via the input unit 21.

[0032] The output unit 22 is an interface unit that is communicatively connected to the touch panel display 15 and the speaker 16. The control unit 18 can transmit a video or image signal to the touch panel display 15 to display it, or transmit an audio signal to the speaker 16 to output sound, via the output unit 22.

[0033] The mass storage device 23 is a non-volatile storage device that is configured with, for example, a hard disk drive, a solid state drive (SSD), a flash memory, etc., and stores various programs such as an operating system and terminal software. The control unit 18 stores, for example, images that it has determined should be stored in the mass storage device 23. Alternatively, the control unit 18 may store, for example, images that it has determined should be stored in an SD card (not shown).

[0034] The speech recognition database 23A (hereinafter referred to as the speech recognition DB 23A) of the mass storage device 23 is a database that stores language models, pronunciation dictionaries, etc. to be used in speech recognition of input speech. When speech is input by the user, the control unit 18 performs speech recognition by referring to the speech recognition DB 23A, and converts the speech into text consisting of character strings.

[0035] Here, an example of a process for setting criteria for images to be stored, which is executed by the control unit 18 of the vehicle-mounted device 10 in the automatic photography system 100, will be described.

[0036] The control unit 18 converts the speech input from the passenger of the vehicle M into text consisting of a character string using the language model of the speech recognition DB 23A described above. Specifically, the control unit 18 recognizes the speech input from the driver DR of the vehicle M, "Take a picture that shows mountains and a lake!", and converts it into text.

[0037] The control unit 18 then identifies and extracts words indicating nouns from the converted text. Specifically, the control unit 18 identifies and extracts the nouns indicating the subject, "mountain" and "lake," from the character string "Take a photo of a mountain and a lake!" converted into text by speech recognition.

[0038] In addition, the control unit 18 may register nouns that are not subjects, such as "photograph," "landscape," and "scenery," as words that will not be used in setting standards for images to be stored, even if they are identified as nouns, and therefore will not be extracted.

[0039] The control unit 18 sets the image containing the words "mountain" and "lake" extracted from the text-converted character string as the image to be stored in the mass storage device 23. In other words, the control unit 18 functions as a standard setting unit that sets the standard for the images to be stored.

[0040] [Configuration of Server Device 11] Next, the configuration of the server device 11 will be described with reference to Fig. 4. Fig. 4 is a block diagram showing an example of the configuration of the server device 11 according to the first embodiment. The server device 11 is a device in which a control unit 26, a communication unit 27, and a large-capacity storage device 28 cooperate with each other via a system bus 25.

[0041] The control unit 26 is configured with a CPU 26A, a ROM 26B, a RAM 26C, etc., and functions as a computer. The CPU 26A reads and executes various programs stored in the ROM 26B and the large-capacity storage device 28, thereby realizing various functions.

[0042] The various programs may be acquired, for example, from another server device or the like via a network, or may be recorded on a recording medium and read via various drive devices. That is, the various programs stored in the mass storage device 28 (including a program for executing processing in the server device 11, which will be described later) can be transmitted via a network, or can be recorded on a computer-readable recording medium and transferred.

[0043] The communication unit 27, like the communication unit 19, is a communication device that transmits and receives data to and from external devices in accordance with instructions from the control unit 26. The communication unit 27 is, for example, a NIC for connecting to a network.

[0044] The control unit 26 receives, for example, data of an image captured by the exterior camera 13 and transmitted from the in-vehicle device 10 via the communication unit 27. The control unit 26 also transmits the caption and aesthetic evaluation value of the image to the in-vehicle device 10 via the communication unit 27.

[0045] The mass storage device 28 is configured by, for example, a hard disk drive, an SSD, a flash memory, or the like, and is a storage device that stores various programs such as an operating system and terminal software.

[0046] The caption generation database 28A (hereinafter referred to as the caption generation DB 28A) of the mass storage device 28 is a database that stores a caption generation model for generating a caption that describes the scene of an image acquired by the control unit 26 via the communication unit 27. The caption generation model is, for example, CoCa (Contrastive Captioner) or BLIP (Bootstrapping Language-Image Pre-training).

[0047] Here, the process of generating a caption for an image performed by the control unit 26 of the server device 11 will be described with reference to Fig. 5. Fig. 5 is a diagram showing an example in which a caption CA indicating the scene of the captured image IM is added to the captured image IM acquired by the control unit 26.

[0048] The control unit 26 inputs, for example, a captured image IM transmitted from the in-vehicle device 10 into a caption generation model stored in the caption generation DB 28A, and obtains as an output a caption CA indicating what is shown in the captured image IM.

[0049] Specifically, the caption generation model recognizes mountain MO and lake LA at the foot of mountain MO as subjects appearing in the captured image IM, as shown in Figure 5, and outputs the caption CA, "mountain and lake at the foot of the mountain."

[0050] In addition, the caption CA of the captured image IM generated by the caption generation model is not limited to being in the form of a sentence, and the output result may be, for example, a group of words consisting of one or more words, such as ``mountain, lake.''

[0051] Referring again to Figure 4, the aesthetic evaluation value calculation database 28B (hereinafter referred to as aesthetic evaluation value calculation DB 28B) of the mass storage device 28 is a database that stores an aesthetic evaluation value calculation model for calculating an aesthetic evaluation value that indicates the beauty of an image acquired by the control unit 26 via the communication unit 27.

[0052] An example of an aesthetic evaluation value calculation model is NIMA (Neural Image Assessment) published by Google. NIMA evaluates, for example, the image quality and overall color balance of an input image, and expresses the evaluation result as an aesthetic evaluation value with a numerical value between 1.0 and 10.0. In NIMA, for example, the more beautiful an image is evaluated, the higher the numerical value assigned.

[0053] An example of the aesthetic evaluation value calculation process performed by the control unit 26 of the server device 11 will now be described with reference to Fig. 6. Fig. 6 is a diagram showing an example in which the control unit 26 calculates the aesthetic evaluation value EV of the captured image IM acquired.

[0054] The control unit 26 inputs the captured image IM transmitted from, for example, the in-vehicle device 10 into the aesthetic evaluation value calculation model stored in the aesthetic evaluation value calculation DB 28B, and obtains as an output result an aesthetic evaluation value EV indicating the beauty of the captured image IM. In other words, the control unit 26 functions as an aesthetic evaluation value calculation unit that calculates the aesthetic evaluation value EV.

[0055] Specifically, as shown in Figure 6, the aesthetic evaluation value calculation model evaluates the image quality and overall color balance of the captured image IM to calculate respective scores, and then multiplies or adds the calculated scores together to output a numerical value such as "7.0 / 10.0" as the final aesthetic evaluation result.

[0056] [Image Storage Determination Process] The following describes the image storage determination process performed by the control unit 18 of the in-vehicle device 10. The control unit 18 of the in-vehicle device 10 acquires the caption CA generated by the server device 11, and evaluates the degree of conformance between the caption CA and the criteria for the image to be stored set by the control unit 18, for example, on a three-level scale of "low," "medium," or "high." In other words, the control unit 18 functions as a conformance calculation unit that calculates the degree of conformance between the criteria for the image to be stored and the caption CA.

[0057] For example, if the criteria for an image to be stored is that it contains "mountain" and "lake," and the caption CA generated by the server device 11 is "mountain and lake at the foot of the mountain," the control unit 18 evaluates the suitability as "high." Note that the manner of evaluating the suitability is not limited to this, and it may be expressed on a five-level scale from 1 to 5, for example.

[0058] Specifically, the control unit 18 evaluates the degree of compatibility based on the degree of agreement between nouns included in the reference (hereinafter also referred to as reference nouns) and nouns included in the caption CA (hereinafter also referred to as caption nouns). More specifically, the control unit 18 evaluates the degree of compatibility based on a numerical value such as "the number of nouns overlapping between the reference noun and the caption noun / the number of nouns included in the caption" itself or on the magnitude of that numerical value.

[0059] In the automatic photography system 100, the control unit 18 acquires the aesthetic evaluation value EV calculated by the server device 11, determines whether the compatibility and aesthetic evaluation value EV are both equal to or greater than a predetermined value, and if the compatibility and aesthetic evaluation value EV are equal to or greater than a predetermined value, stores the image used in the determination. In other words, the control unit 18 functions as a determination unit that determines whether an image should be stored.

[0060] When the judgment results for the compatibility and aesthetic evaluation value EV are, for example, "high" and "6.5" or higher, the control unit 18 stores the image used for the judgment in the mass storage device 23. That is, the control unit 18 functions as a storage control unit that stores the image in the mass storage device 23 serving as a storage unit.

[0061] For example, in response to an image call operation by the user of the in-vehicle device 10, the control unit 18 displays the image stored in the mass storage device 23 on the touch panel display 15. This allows the user of the in-vehicle device 10 to obtain an image that is automatically taken while traveling in the vehicle M and that the user desires, i.e., an image that matches the user's preferences.

[0062] 7 and 8, the specific operation of the in-vehicle device 10 in this embodiment will be described. FIG. 7 is a flowchart showing a reference setting routine RT1 executed by the control unit 18 of the in-vehicle device 10.

[0063] The control unit 18 starts the reference setting routine RT1, for example, when power is turned on to the in-vehicle device 10. Note that the reference setting routine RT1 is repeatedly executed by the control unit 18 as long as the in-vehicle device 10 is powered on.

[0064] First, the control unit 18 determines whether a reference setting operation has been received (step S101). For example, the control unit 18 determines that an operation for starting the reference setting process (reference setting start operation) has been performed when a user (e.g., the driver DR of the vehicle M) calls the wake word of the in-vehicle device 10 or makes a predetermined voice input such as "Make a note!" If the control unit 18 determines that a reference setting start operation has not been received (step S101: NO), it ends the reference setting routine RT1.

[0065] When the control unit 18 determines that the reference setting start operation has been received (step S101: YES), the control unit 18 receives voice input from the user (step S102). For example, as described above, the control unit 18 acquires voice input from the driver DR of the vehicle M via the microphone 14. Note that the control unit 18 may end the reference setting routine RT1 if there is no voice input from the user within a certain period of time in step S102.

[0066] After step S102, the control unit 18 performs speech recognition on the input speech using the speech recognition DB 23A of the mass storage device 23, and converts the speech into text (step S103). For example, the control unit 18 converts a speech input by the user, such as "Take a picture of a mountain and a lake!", into text consisting of a character string.

[0067] After step S103, the control unit 18 extracts a noun from the text as the speech recognition result, and sets an image including a subject indicated by the noun as a criterion for an image to be stored (step S104). For example, the control unit 18 sets an image including "mountain" and "lake" in the speech as a criterion for an image to be stored. After step S104, the control unit 18 ends the criterion setting routine RT1.

[0068] 8 is a flowchart showing an image storage routine RT2 executed by the control unit 18 of the in-vehicle device 10. The control unit 18 starts the image storage routine RT2, for example, when the power supply to the in-vehicle device 10 is turned on. Note that the image storage routine RT2 is repeatedly executed by the control unit 18 as long as the power supply to the in-vehicle device 10 is turned on.

[0069] First, the control unit 18 determines whether or not a reference for the image to be stored has been set (step S201). For example, the control unit 18 determines whether or not the reference has been set by the control unit 18 in response to voice input from the user in the reference setting routine RT1. If the control unit 18 determines that the reference has not been set (step S201: NO), the image storage routine RT2 is terminated.

[0070] When the control unit 18 determines that a criterion has been set (step S201: YES), the control unit 18 acquires a caption that indicates the scene depicted by the image and that has been generated by the control unit 26 of the server device 11 (step S202). For example, the control unit 18 acquires the caption CA, "Mountains and a lake at the foot of the mountains," generated using the above-described caption generation model CoCa.

[0071] After step S202, the control unit 18 calculates the degree of conformance between the criteria set in the criteria setting routine RT1 and the caption acquired in step S202 (step S203). Specifically, the control unit 18 evaluates the degree of conformance between the criteria of the image to be stored and the caption CA as described above, for example, on a three-level scale of "low," "medium," and "high."

[0072] After step S203, the control unit 18 acquires an aesthetic evaluation value indicating the beauty of the image calculated by the control unit 26 of the server device 11 (step S204). Specifically, the control unit 18 acquires an aesthetic evaluation value EV of 1.0 to 10.0 calculated using, for example, the above-mentioned aesthetic evaluation value calculation model, NIMA.

[0073] After step S204, the control unit 18 determines whether the degree of conformance calculated in step S203 and the aesthetic evaluation value EV acquired in step S204 are both equal to or greater than a predetermined value (step S205). Specifically, the control unit 18 determines whether the degree of conformance is "high" and the aesthetic evaluation value EV is "6.5" or greater, for example.

[0074] If the control unit 18 determines that neither the compatibility nor the aesthetic evaluation value EV is equal to or greater than a predetermined value, or if only one of them is equal to or greater than a predetermined value (step S205: NO), the control unit 18 ends the image storage routine RT2.

[0075] When the control unit 18 determines that both the compatibility and aesthetic evaluation value EV are above a predetermined level, for example, as described above, the compatibility is “high” or above and the aesthetic evaluation value EV is “6.5” or above (step S205: YES), the control unit 18 stores the image used in the determination in the large-capacity storage device 23.

[0076] According to the above-described standard setting routine RT1 and image storage routine RT2, landscape images captured by the exterior camera 13 are automatically stored while the vehicle M is traveling, based on the standard for images to be stored that is set in response to voice input by the user of the in-vehicle device 10. Therefore, according to the automatic photography system 100 of this embodiment, the user of the in-vehicle device 10 can obtain images that are automatically captured according to the user's wishes (preferences) while traveling in the vehicle M.

[0077] Furthermore, in the automatic photography system 100 of this embodiment, images with an aesthetic evaluation value EV, which indicates the beauty of the image, equal to or greater than a predetermined value are stored. In other words, images with extremely low image quality or images containing subjects that detract from the beauty of the surrounding scenery are not stored. Therefore, the user of the in-vehicle device 10 can obtain images that meet his or her preferences and that have a certain level of beauty or greater.

[0078] In this embodiment, the control unit 18 of the in-vehicle device 10 stores an image when both the degree of conformance between the caption and the standard and the aesthetic evaluation value of the image are equal to or greater than a predetermined value. However, this is not limited to this. For example, the control unit 18 may store an image based only on the degree of conformance between the caption and the standard. In other words, the control unit 26 of the server device 11 does not need to calculate the aesthetic evaluation value of the image.

[0079] In this embodiment, the control unit 18 recognizes a noun from the user's input speech, but it may also identify a modifier that modifies the noun in addition to the noun. For example, if the recognition result of the input speech shows that the noun "lake" is accompanied by a positive modifier such as "beautiful," the threshold for the aesthetic evaluation value of the image may be increased compared to when no positive modifier is attached or when a negative modifier is attached.

[0080] In this embodiment, if the voice input by the user includes words indicating a season, date, or time zone, the control unit 18 may set criteria for the images to be stored based on the content indicated by the words indicating the season, date, or time zone.

[0081] For example, if the voice input by the user contains the word "night," the control unit 18 may set it as a criterion for storing images taken at a time period after 19:00.

[0082] Furthermore, if the voice input by the user contains the word "weekend," the control unit 18 may set images captured on Saturday or Sunday as the standard for storage. In this case, the control unit 18 may calculate the degree of conformance between the image captured by the exterior camera 13 and the standard only when the current date and time corresponds to the season, date, or time period set as the standard for images to be stored.

[0083] Furthermore, in this embodiment, if the voice input by the user includes words that identify the shooting location, the control unit 18 may set the words as the reference for the image to be stored. For example, if the voice input by the user includes the word "Tokyo," the control unit 18 may set the images to be stored based on the current location of the vehicle M, which are captured while the vehicle M is in "Tokyo." In this case, the control unit 18 may calculate the degree of conformity between the image captured by the exterior camera 13 and the reference only when the current location of the vehicle M corresponds to the location set as the reference for the image to be stored.

[0084] In this embodiment, the control unit 18 sets the criteria for the images to be stored based on the subject recognized by voice input from the user, but this is not limited to this, and the criteria for the images to be stored may be set based on character input by the user via the touch panel display 15 instead of voice input via the microphone 14.

[0085] Furthermore, in this embodiment, the control unit 18 of the in-vehicle device 10 acquires the caption of the image and the aesthetic evaluation value of the image from the server device 11, but this is not limiting, and all processing may be performed on the side of the in-vehicle device 10. That is, the control unit 18 of the in-vehicle device 10 may function as the image acquisition unit, the standard setting unit, the compatibility calculation unit, the aesthetic evaluation value calculation unit, the determination unit, and the storage control unit.

[0086] In addition, in this embodiment, the control unit 26 of the server device 11 generates a caption for the captured image transmitted from the in-vehicle device 10 and calculates an aesthetic evaluation value of the captured image, but this is not limiting, and some or all of other processing may be performed on the server device 11 side. Specifically, the server device 11 may acquire voice data of a voice input to the microphone 14 via the communication unit 27 and perform voice recognition processing to set criteria for images to be stored.

[0087] The server device 11 may also acquire a captured image transmitted from the in-vehicle device 10, determine whether the compatibility and aesthetic evaluation value EV of the acquired captured image are both equal to or greater than a predetermined value, and, if it determines that both are equal to or greater than a predetermined value, store the image used for the determination in the mass storage device 28. That is, the control unit 26 of the server device 11 may function as at least one of the image acquisition unit, the standard setting unit, the compatibility calculation unit, the aesthetic evaluation value calculation unit, the determination unit, and the storage control unit. In this case, the control unit 18 of the in-vehicle device 10 may be configured to execute functions not executed by the server device 11.

[0088] [Variation 1] Variations 1 to 3 of the automatic photography system 100 according to the first embodiment will be described below. First, Variation 1 of the first embodiment will be described with reference to Fig. 9. Variation 1 differs from the first embodiment in that weighting is applied to the aesthetic evaluation value calculated by the control unit 26 of the server device 11, but is otherwise similar to the first embodiment.

[0089] 9 is a table TB1 showing the weighting for each subject stored in the mass storage device 28. In this modification, the control unit 26 of the server device 11 communicates with a device having an imaging function owned by the user of the in-vehicle device 10, such as a smartphone or a digital camera, to acquire images stored in the device.

[0090] The control unit 26 recognizes subjects appearing in images acquired from a device owned by the user of the in-vehicle device 10 using a known object recognition technology such as YOLO, and classifies the images by the recognized subjects. For example, as shown in FIG. 9, if an image shows the sea, the image is considered to be an image corresponding to "sea" and the term "number of images" is incremented by 1. Also, for example, if an image shows a temple and a cat, the terms "number of images" for "temple" and "cat" are each incremented by 1.

[0091] The control unit 26 classifies the acquired images by subject, for example, and creates table TB1 by assigning a large weight to the subjects with the most images, in other words, the subjects with the most frequent appearance, as the user's preference. For example, as shown in Fig. 9, the control unit 26 assigns a weight to "sea," which has the most images, to "1.5," and assigns a weight to "temple," which has a relatively small number of images, to "0.3."

[0092] In this modification, the control unit 26 corrects the aesthetic evaluation value using the weight values ​​shown in Table TB1 in step S204 of the image storage routine RT2 described above. Specifically, after calculating the aesthetic evaluation value of the image, the control unit 26 multiplies the aesthetic evaluation value by the weight value corresponding to the subject appearing in the image based on Table TB1 of Fig. 9. For example, if the image includes an "ocean," the control unit 26 multiplies the aesthetic evaluation value by "1.5" based on Table TB1.

[0093] It should be noted that the correction of the aesthetic evaluation value using the weighting values ​​in Table TB1 does not have to be applied to all of the subjects appearing in the image; for example, only the subject that occupies the largest area in the image or the subject appearing in the center of the image may be subject to correction.

[0094] In this modification, the control unit 18 of the in-vehicle device 10 acquires the aesthetic evaluation value corrected by the control unit 26, and determines whether or not to store the image based on the acquired corrected aesthetic evaluation value.

[0095] According to this modified automatic photography system 100, by correcting the aesthetic evaluation value using the weighting values ​​in Table TB1, it becomes easier for the user of the in-vehicle device 10 to memorize images that contain subjects that the user has photographed many times in the past.

[0096] Furthermore, with the automatic photography system 100 of this modified example, the user's preferences are more likely to be reflected even if the user's input voice does not contain a noun that indicates a subject that the user has photographed many times in the past. Therefore, with the automatic photography system 100 of this modified example, the user of the in-vehicle device 10 can obtain images that are automatically photographed according to the user's preferences while traveling in the vehicle M.

[0097] In this modified example, the control unit 26 corrects the aesthetic evaluation value by referring to table TB1 stored in the large-capacity storage device 28. However, the storage destination of table TB1 is not limited to this. For example, a database server capable of storing and updating table TB1 may be provided separately from the server device 11, and the control unit 26 of the server device 11 may acquire table TB1 from the database server. Alternatively, a mobile device such as a smartphone that stores a large number of images taken by a user in the past may store an application capable of calculating, storing, updating, and communicating table TB1, and the application may be executed on the mobile device to transmit table TB1 to the in-vehicle device 10 or the server device 11 via the respective communication units.

[0098] Furthermore, in this modified example, the control unit 26 determines the weight value according to the number of images for each subject, but this is not limited to this. For example, the weight value may be calculated by taking into account the shooting location information and shooting date and time of the images stored in the device of the user of the in-vehicle device 10.

[0099] For example, if the images stored in the user's device are often taken in the morning, and the date and time of the image for which the aesthetic evaluation value is calculated is in the morning, the control unit 26 may correct the calculated aesthetic evaluation value using a weight value set higher.

[0100] Furthermore, in addition to or instead of setting weight values ​​for each subject of the images stored in the user's device, the control unit 26 may set weight values ​​for each scene of the image, such as a natural landscape, an urban landscape, etc. Note that the weight values ​​for each subject and each scene may be configured to be set by the user himself / herself.

[0101] Furthermore, in this embodiment, the server device 11 corrects the aesthetic evaluation value using the weight value that reflects the user's preference, but this is not limited thereto, and the control unit 18 of the in-vehicle device 10 may calculate the aesthetic evaluation value. In this case, the table TB1 that reflects the user's preference is generated by the control unit 26 of the server device 11, and the control unit 18 of the in-vehicle device 10 may request the weight value for each subject shown in table TB1 from the server device 11 via the communication unit 19 at any timing, and calculate a corrected aesthetic evaluation value based on the weight value for each subject obtained from the server device 11.

[0102] Next, Modification 2 of Example 1 will be described with reference to Fig. 10. Modification 2 differs from Example 1 in the manner in which the control unit 26 of the server device 11 learns the aesthetic evaluation value of the model when calculating the aesthetic evaluation value, but is otherwise similar to Example 1.

[0103] 10 is a diagram showing a learning mode of the aesthetic evaluation value. In this modification, the control unit 26 collects an unspecified number of posted images that are posted together with comments on SNS (for example, X or Instagram), and causes the aesthetic evaluation value calculation model to learn the aesthetic evaluation value of the posted image, which is calculated based on the content of the comments, by linking the posted image to the aesthetic evaluation value.

[0104] Specifically, when the control unit 26 acquires a posted image from a social networking site, it evaluates whether or not the comments attached to the posted image are positive, and to what extent, using, for example, a text emotion recognition model that estimates emotions from text.

[0105] For example, as shown in Figure 10, if an image posted on SNS is accompanied by a comment such as "Mount Fuji seen from Lake Motosu! It was so beautiful!", the control unit 26 sets the aesthetic evaluation value of the posted image higher than the aesthetic evaluation value in the absence of the comment, and re-trains the aesthetic evaluation value calculation model using a data group including multiple pieces of data, each consisting of the set aesthetic evaluation value and the posted image.

[0106] According to the server device 11 of this modified example, an unspecified number of posted images that are posted to SNS along with comments can be collected, and the aesthetic evaluation value of the posted image calculated based on the content of the comments can be linked to the posted image and learned by the aesthetic evaluation value calculation model, thereby allowing the aesthetic evaluation value calculation model to calculate an aesthetic evaluation value that is closer to human aesthetic sense.

[0107] Therefore, according to the automatic photography system 100 of this modified example, by using an aesthetic evaluation value calculation model that better reflects people's aesthetic sense, it is possible to prevent differences from occurring between the aesthetic evaluation value calculated by the aesthetic evaluation value calculation model and the aesthetic sense of the majority of people.

[0108] Note that the collection of posted images and comments attached to posted images posted to the SNS as shown in this modified example may be performed by a server device other than the server device 11. In other words, instead of the server device 11 itself collecting posted images and the like and generating learning data, another server may generate learning data for retraining the model in the server device 11, and the server device 11 may simply receive the data.

[0109] Furthermore, in this embodiment, the server device 11 calculates the aesthetic evaluation value using an aesthetic evaluation value calculation model that has been retrained based on images posted by an unspecified number of people and the contents of comments attached to the posted images, but this is not limited to this, and the aesthetic evaluation value calculation model may be provided on the in-vehicle device 10 side, and the control unit 18 of the in-vehicle device 10 may calculate the aesthetic evaluation value. In this case, the re-learning of the aesthetic evaluation value calculation model is performed on the server device 11 side, and the control unit 18 of the in-vehicle device 10 may request update information on the re-trained aesthetic evaluation value calculation model from the server device 11 via the communication unit 19 at any timing, and update the conventional aesthetic evaluation value calculation model based on the update information acquired from the server device 11.

[0110] Next, a third modification of the first embodiment will be described with reference to Fig. 11. The third modification is different from the first embodiment in the processing performed after the control unit 18 stores the image, but is otherwise similar to the first embodiment.

[0111] 11 is a flowchart showing a reference correction routine RT3 executed by the control unit 18 of the in-vehicle device 10. The control unit 18 starts the reference correction routine RT3, for example, when the in-vehicle device 10 is powered on. Note that the reference correction routine RT3 is repeatedly executed by the control unit 18 as long as the in-vehicle device 10 is powered on.

[0112] First, the control unit 18 determines whether or not an image has been stored in the image storage routine RT2 (step S301). If the control unit 18 determines that an image has not been stored (step S301: NO), the control unit 18 ends the reference correction routine RT3.

[0113] When the control unit 18 determines that the image has been stored (step S301: YES), it displays the stored image on the touch panel display 15 (step S302). Furthermore, the control unit 18, for example, displays the image on the touch panel display 15 and outputs a voice message via the speaker 16 asking the user of the in-vehicle device 10 whether to modify the reference. Note that the image may be displayed on the touch panel display 15 each time an image is stored, or may be displayed when a predetermined number of images have been stored.

[0114] After step S302, the control unit 18 determines whether or not a request to modify the criteria has been received from the user (step S303). If the control unit 18 has not received a request to modify the criteria from the user (step S303: NO), for example, if the control unit 18 receives a negative response such as "no modification," or if there is no voice input from the user for a certain period of time, the control unit 18 determines that the modification request has not been received and ends the criteria modification routine RT3.

[0115] If the control unit 18 determines that it has received a request from the user to modify the criteria (step S303: YES), it modifies the criteria set in the criteria setting routine RT1 and sets the modified criteria as the new criteria (step S304).

[0116] Specifically, when the user checks an image stored in the mass storage device 23 on the touch panel display 15 and places an additional order, for example, if the user gives a voice input such as "please make it one that also includes cherry blossoms," the control unit 18 modifies the image to include "cherry blossoms" in addition to "mountains" and "lakes" as the standard for the new image to be stored.

[0117] The automatic photography system 100 of this modified example can flexibly accommodate user requests, such as by changing some of the criteria when an image that the user likes is acquired. Therefore, the automatic photography system 100 of this modified example can store images that better match the user's preferences.

[0118] In this modified example, the control unit 18 may generate an image image based on the revised criteria, for example, using an image generation model, and display the generated image image on the touch panel display 15 as an image for the user to confirm.

[0119] For example, when the control unit 18 receives an order from a user specifying "cherry blossoms" in addition to "mountains" and "lakes" as criteria for images to be memorized, the control unit 18 may output and display an image in which cherry blossom trees are combined with an image of "mountains" and "lakes" in the image generation model.

[0120] Next, an automatic photography system 100 according to a second embodiment will be described with reference to Fig. 12. The second embodiment is similar to the first embodiment except for the manner in which the criteria are set in the in-vehicle device 10 according to the content of the user's speech. Fig. 12 is a diagram showing the configuration of the front seat area of ​​the vehicle M and the interior of the vehicle M, similar to Fig. 2.

[0121] In this embodiment, the vehicle M is also assumed to pass through a place with abundant nature, such as mountains and a lake, while traveling from the departure point of the vehicle M to the destination. Fig. 12 shows an example of a user making a statement to the in-vehicle device 10, in which the driver DR of the vehicle M wants to capture a landscape photo that captures a sense of nature while traveling in the vehicle M, and utters the words "Take a photo of a simple landscape!" into the microphone 14.

[0122] In this embodiment, when the voice uttered by the user does not contain any nouns that identify the subject, but only modifiers that describe the atmosphere of the photograph or landscape, the control unit 18 of the in-vehicle device 10 identifies the modifiers contained in the recognized text and sets the scene represented by the identified modifiers as the standard to be memorized.

[0123] Specifically, the control unit 18 identifies adjectives, adjectival verbs, and adverbs related to terms such as "photo," "atmosphere," and "scenery" contained in speech such as "I want a photo with a relaxing atmosphere," "Take a photo of a rustic landscape," "I want a romantic photo," or "Take a photo with a lively atmosphere" uttered by the user.

[0124] For example, when the control unit 18 identifies the adjective "simple" as a modifier related to "landscape," it sets an image of a landscape with houses standing in the middle of a vast field as the standard image to be memorized, and sets images including "mountains," "fields," "houses," etc.

[0125] In this way, even if the voice uttered by the user does not contain a noun that identifies the subject, the control unit 18 can set criteria for the image to be stored by identifying adjectives, adjectival verbs, and adverbs related to terms such as "photograph" and "landscape."

[0126] The control unit 18 may use images with comments (atmosphere evaluations) such as “relaxed,” “simple,” “romantic,” “bright,” and “lively,” obtained from SNS, as training data, and set the conditions for the images to be stored using a model that has learned the correspondence between the subjects in the images and the atmosphere evaluations.

[0127] Next, an automatic photography system 100 according to a third embodiment will be described with reference to Fig. 13 and Fig. 14. The third embodiment is similar to the first embodiment in other respects, except for the processing performed on the aesthetic evaluation value calculated by the server device 11.

[0128] FIG. 13 shows an example of a waveform EW (upper part of the figure) of a time-series aesthetic evaluation value calculated by the control unit 26 of the server device 11, and a trend waveform TW (lower part of the figure) obtained by smoothing the waveform.

[0129] In the automatic photography system 100 of this embodiment, the control unit 26 of the server device 11 sequentially calculates the aesthetic evaluation values ​​of the images captured by the outside camera 13 and transmitted from the control unit 18 of the in-vehicle device 10, i.e., the temporally consecutive captured images, and obtains a waveform EW, which is the time series value of the aesthetic evaluation values ​​of the temporally consecutive captured images.

[0130] Furthermore, the control unit 26 of the server device 11 performs noise removal on the obtained waveform EW using a signal processing method such as empirical mode decomposition (EMD) to generate a trend waveform TW, which is an evaluation value curve obtained by smoothing the waveform EW. In other words, the control unit 26 functions as a smoothed evaluation value generating unit that generates a smoothed evaluation value obtained by smoothing temporally continuous aesthetic evaluation values.

[0131] In the automatic photography system 100 of this embodiment, the control unit 18 of the in-vehicle device 10 acquires the trend waveform TW generated by the control unit 26 of the server device 11, identifies a captured image corresponding to the timing at which the maximum value MA appears in the acquired trend waveform TW, and stores the identified captured image in the mass storage device 23. In other words, the control unit 18 functions as an identification unit that identifies an image corresponding to the timing at which the smoothing evaluation value satisfies a predetermined criterion.

[0132] For example, when the aesthetic evaluation value calculation model shown in Example 1 is used to calculate the aesthetic evaluation value for each frame of the image captured by the exterior camera 13, depending on the influence of sunlight, etc., the aesthetic evaluation value may change significantly between adjacent frames even if the same scene is captured by the exterior camera 13.

[0133] If such a change in aesthetic evaluation value occurs even when almost the same scene is photographed by the outside-vehicle camera 13, when calculating the aesthetic evaluation value of images acquired at a certain time interval as in Example 1, there is a risk that the image of that scene will not be stored because the aesthetic evaluation value just so happens to be low for the moment when it is calculated. In other words, even if the scenery is one that the user likes, there is a risk that the image will not be recorded because the aesthetic evaluation value was momentarily low.

[0134] According to the in-vehicle device 10 of this embodiment, captured images corresponding to the timing of the maximum value MA appearing in the trend waveform TW generated by the control unit 26 of the server device 11 are extracted and the extracted captured images are stored, thereby recording all images with particularly high aesthetic evaluation values ​​within a certain time range. Therefore, according to the in-vehicle device 10 of this embodiment, it is possible to prevent images that meet the user's preferences from being omitted from the recording.

[0135] In this embodiment, it is not necessary to perform the evaluation of the compatibility between the caption and the standard by the server device 11 described in the first embodiment. That is, the server device 11 of this embodiment may only calculate the aesthetic evaluation value and generate the trend waveform TW.

[0136] In this way, when the server device 11 only calculates the aesthetic evaluation value and generates the trend waveform TW, the control unit 18 of the in-vehicle device 10 can acquire all images of the aesthetic evaluation value corresponding to the maximum value of the trend waveform TW from the server device 11, and then present the acquired images to the user, and the user can select his or her favorite image from the presented images.

[0137] 14, the control unit 18 displays images of aesthetic evaluation values ​​corresponding to the maximum values ​​of the trend waveform TW side by side in small thumbnail form at the bottom of the touch panel display 15, and displays two images arbitrarily selected by the user from these images in a large, comparable form at the center of the touch panel display 15. In other words, the control unit 18 functions as a display control unit that displays a plurality of images in a comparable form on the touch panel display 15 as a display unit.

[0138] The user of the in-vehicle device 10 can select images to keep in the mass storage device 23, for example, by checking a check box provided for each image displayed on the touch panel display 15. For example, images for which the check box is not checked are deleted from the mass storage device 23.

[0139] By allowing the user to select the images to be kept later, the user can decide for himself or herself whether or not to keep the images. Therefore, after obtaining images that are automatically taken while traveling in the vehicle M, the user of the in-vehicle device 10 can select the images to be kept according to his or her preferences.

[0140] Note that the image comparison mode shown in Figure 14 is just one example, and it is also possible to have the user select which image to keep by, for example, displaying one image selected by the user in the center of the touch panel display 15 for a few seconds and then switching to another image, i.e., by switching the images one by one every few seconds.

[0141] In this embodiment, the control unit 18 stores images at timings corresponding to the maximum values ​​of the trend waveform TW of the aesthetic evaluation value, but may also store images at timings corresponding to the minimum values. Images showing minimum aesthetic evaluation values ​​may not be beautiful, but may be enjoyable for the user to look back on later or may become a hot topic on social media. Therefore, the user may be allowed to select whether or not to store the images.

[0142] Next, a fourth modification of the automatic photography system 100 of the third embodiment will be described with reference to Fig. 15. This modification differs from the third embodiment in the recording mode of the image of the trend waveform TW, but is otherwise similar to the third embodiment.

[0143] 15 is a diagram showing the trend waveform TW in this modified example. In this modified example, as shown in FIG. 15, the control unit 18 sets a threshold value TH for the trend waveform TW acquired from the server device 11, and stores captured images corresponding to all timings during a period in which the threshold value TH is exceeded as images with high aesthetic evaluation.

[0144] According to the in-vehicle device 10 of this modification, captured images corresponding to a period in which the trend waveform TW generated by the control unit 26 of the server device 11 exceeds the threshold value TH are identified, and the identified captured images are stored, thereby recording all images with particularly high aesthetic evaluation values ​​within a certain time range. Therefore, according to the in-vehicle device 10 of this modification, it is possible to prevent images that meet the user's preferences from being omitted from the recording.

[0145] In the above-described third embodiment and fourth modification, the control unit 26 generates the trend waveform TW using EMD, but the method for smoothing (removing noise from) the waveform EW of the aesthetic evaluation value is not limited to this. For example, instead of the empirical mode decomposition, a processing method such as a Kalman filter, a median filter, a moving average filter, or a wavelet transform may be adopted.

[0146] Furthermore, the configurations according to the above-described embodiments and modifications can be combined in any manner within a range that does not cause contradictions.

[0147] 100 Automatic photography system device 10 In-vehicle device 11 Server device 13 Outside-vehicle camera 14 Microphone 15 Touch panel display 16 Speaker 18, 26 Control unit 19, 27 Communication unit 21 Input unit 22 Output unit 23, 26 Large-capacity storage device

Claims

1. An information processing system comprising: an image acquisition unit that acquires images captured by an imaging unit that moves with a moving object; a standard setting unit that sets standards for images to be stored based on language input by a user; a compatibility calculation unit that calculates the compatibility of one image acquired by the image acquisition unit with the standards; a judgment unit that determines whether or not the one image should be stored based on the compatibility of the one image; and a memory control unit that stores images that the judgment unit determines should be stored in a memory unit.

2. The information processing system according to claim 1, wherein the criterion setting unit sets the criterion as including a subject represented by a noun contained in a natural language.

3. The information processing system according to claim 1, wherein the criterion setting unit sets the criterion as including a scene expressed by a modifier contained in a natural language.

4. The information processing system described in claim 1, characterized in that the compatibility calculation unit generates a sentence indicating the scene depicted by the one image, or a group of words containing one or more words, and calculates the compatibility by comparing the generated sentence or group of words with the criteria.

5. The information processing system described in claim 1, characterized in that the criteria setting unit modifies the criteria based on the language input by the user after the memory control unit stores the image that the judgment unit has determined should be stored in the memory unit.

6. An information processing system as described in any one of claims 1 to 5, characterized in that it has an aesthetic evaluation value calculation unit that calculates an aesthetic evaluation value that indicates the beauty of the one image, and the judgment unit judges whether or not the one image should be stored based on the aesthetic evaluation value of the one image.

7. The information processing system described in claim 6, characterized in that the aesthetic evaluation value calculation unit corrects the aesthetic evaluation value of the one image using a weighting value for each subject or scene set based on the user's operation or the user's preference, and the judgment unit judges whether or not the one image should be stored based on the corrected aesthetic evaluation value.

8. The information processing system described in claim 6, characterized in that the aesthetic evaluation value calculation unit calculates the aesthetic evaluation value of the one image using a model that has been trained by linking the aesthetic evaluation value of the posted image, which is calculated based on the content of posted images posted by an unspecified number of people and the comments attached to the posted images, with the posted image.

9. An information processing system according to claim 6, wherein said standard setting section sets at least one of season, date, time zone, and shooting location as the standard for the images to be stored.

10. An information processing method executed by an information processing system, comprising: an image acquisition step for acquiring images captured by an imaging unit moving together with a moving object; a standard setting step for setting standards for images to be stored based on language input by a user; a suitability calculation step for calculating the suitability of one image acquired in the image acquisition step with respect to the standards; a determination step for determining whether or not the one image should be stored based on the suitability of the one image; and a storage control step for storing images determined to be stored in the determination step in a storage unit.

11. A program to be executed by a computer, comprising: an image acquisition step for acquiring images captured by an imaging unit moving with a moving object; a standard setting step for setting standards for images to be stored based on language input by a user; a suitability calculation step for calculating the suitability of one image acquired in the image acquisition step with respect to the standards; a determination step for determining whether or not the one image should be stored based on the suitability of the one image; and a storage control step for storing images determined to be stored in the determination step in a storage unit.

12. A recording medium having recorded thereon a program that causes a computer provided in an information processing system to execute the following steps: an image acquisition step for acquiring images captured by an imaging unit that moves with a moving object; a standard setting step for setting standards for images to be stored based on language input by a user; a suitability calculation step for calculating the suitability of one image acquired in the image acquisition step with respect to the standards; a determination step for determining whether or not the one image should be stored based on the suitability of the one image; and a storage control step for storing images determined to be stored in the determination step in a storage unit.

Images