Imaging equipment, control methods for imaging equipment, programs

Abstract

To solve the problem in difficulty efficiently recording an operation history in some cases.SOLUTION: When a user changes the orientation of an imaging apparatus by a predetermined amount or more, when the user moves the imaging apparatus by a predetermined amount or more, and when the user gives a playback operation instruction to an imaging operation apparatus, the previous operation history is deleted and not stored as an imaging history.SELECTED DRAWING: Figure 6

Description

Technical Field

[0006] , , , ,

[0001] The present invention relates to a device having a photographing function, and particularly to automatic photographing.

Background Art

[0002] In recent years, digital cameras and smartphones, which are mobile devices equipped with imaging devices, are configured to perform photographing while displaying a live view image on a display device such as an LCD.

[0003] Although the user is performing live view display, since the user is not always conscious of taking a photograph, there is a case where a scene preferred by the user is being displayed in the live view but the opportunity to take a photograph is missed.

[0004] Therefore, devices have been devised that automatically discriminate scenes according to the user's preferences using AI or the like and perform photographing. Analytical data for discriminating the user's preferred images is required. In terms of AI, this is learning data for creating an inference model. Therefore, in Patent Document 1, a proposal has been made to store all operations performed by the user for the purpose of analytical data and learning for discriminating the user's preferred images.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] However, the prior art disclosed in the aforementioned patent documents has the problem that if all user operations are memorized, the operation data becomes enormous. Furthermore, if all user operations are memorized, a lot of data that is invalid for determining user preferences is included. Therefore, the present invention aims to record operation history more efficiently. [Means for solving the problem]

[0007] To achieve the above objective, the present invention To obtain an image First Imaging means, A second imaging means for acquiring an image, Input section for the user to operate the device. And, It comprises an image unit and a storage unit that stores the captured image and the operation history of the input unit in association with each other. If the user performs a selection operation from the input unit to the imaging device to switch between using the first imaging means and using the second imaging means, the selection operation This method is characterized by deleting previous operation history and storing it in the memory unit. [Effects of the Invention]

[0008] According to the present invention, operation history can be recorded more efficiently. [Brief explanation of the drawing]

[0009] [Figure 1] This is a system diagram of the first embodiment. [Figure 2] This is a hardware resource diagram of each device in the first embodiment. [Figure 3] This is a software configuration diagram of the first embodiment. [Figure 4] This is a diagram of the learning model according to the first embodiment. [Figure 5] This is a diagram illustrating the operation of the system according to the first embodiment. [Figure 6] This is a flowchart of the operation of the first embodiment. [Figure 7] This is a flowchart of the operation of the first embodiment. [Figure 8] This is a flowchart of the operation of the first embodiment. [Figure 9]It is a data structure diagram of the first embodiment. [Figure 10] It is a table data structure diagram of the first embodiment.

Embodiment for Carrying Out the Invention

[0010] The present invention relates to an imaging device for creating a learning data for creating an inference model for automatically determining whether a live image is in line with a user's preference when the imaging device is operating in a live view mode, and for automatically taking a picture using the inference model.

[0011] The system to which the present invention can be applied will be described with reference to FIG. 1.

[0012] 101 is a digital camera used by a user.

[0013] 102 is a client terminal capable of connecting to the digital camera 101 and transmitting and receiving data.

[0014] 103 is a communication means such as USB that enables communication between the digital camera 101 and the client terminal 102.

[0015] As an example of the communication means 103, a wired communication means such as UBS may be used, or a wireless communication means such as WiFi or Bluetooth may also be used.

[0016] 104 is a learning server capable of learning a deep learning model.

[0017] 105 is a data server.

[0018] 106 connects the client terminal 102, the learning server 104 and the data server 105 to enable communication.

[0019] The hardware resources of each device constituting the system of FIG. 1 will be described with reference to FIG. 2.

[0020] Unit 201 is the CPU for overall control of the digital camera 101, and also serves as the power supply control unit.

[0021] 202 is a ROM that holds the programs and data necessary for CPU201 to operate.

[0022] 203 is RAM used to temporarily expand the program read by CPU201 from ROM202, execute the program from there, and temporarily hold operational data.

[0023] The 204 is a GPU specifically designed for data processing, enabling high-speed image processing and matrix operations, and the processing of larger amounts of data.

[0024] GPU204 can perform calculations efficiently by processing more data in parallel, so it is effective to use GPU204 when performing inference with a pre-trained model trained using deep learning.

[0025] Therefore, in the digital camera 101, when performing inference processing using a pre-trained model trained with deep learning, the GPU 204 is used in addition to the CPU 201.

[0026] Specifically, when performing estimation processing using a pre-trained model, CPU201 and GPU204 work together to perform calculations.

[0027] Furthermore, the estimation process may be performed using only CPU201 or GPU204.

[0028] 205 is a movement detection unit consisting of a gyro sensor, an accelerometer, and other components for detecting changes in the orientation or movement of the digital camera 101.

[0029] 206 is an I / F unit for exchanging data between the digital camera 101 and the client terminal 102 via the communication means 103.

[0030] 207 is an operation unit that allows the user to perform shooting and playback operations, menu operations, etc. on the digital camera 101, and an input unit that receives user input such as detecting when the user looks through the camera or giving instructions to the digital camera 101 based on the user's gaze.

[0031] 208 is a display unit that displays images being taken or taken by the rear camera 212 and front camera 213, which will be described later, and also displays the operating status of the cameras.

[0032] 209 is a camera engine that processes images acquired by the rear camera 212 and the front camera 213, and also performs image processing for displaying images stored in the storage unit 210 (described later) on the display unit 208.

[0033] 210 is a storage unit for saving still images and video image data captured by the digital camera 101. 211 is a system bus for connecting 201 to 214, which are the various blocks that make up the digital camera 101. 212 is an out-camera consisting of an image sensor for taking pictures. 213 is an in-camera consisting of an image sensor for taking pictures, which can capture images of the user on the opposite side from the out-camera 212.

[0034] 214 is a timer unit that measures the system time of the digital camera 101 and the time of day. 215 is a CPU that controls the entire client terminal 102.

[0035] 216 is an HDD that holds programs and data necessary for the CPU 215 to operate. 217 is RAM that temporarily expands the programs read by the CPU 215 from HDD 216, executes the programs from there, and also temporarily holds operational data. 218 is a NIC that communicates with the learning server 104 and data server 105 via the local network 106.

[0036] 219 is an interface unit for exchanging data between the client terminal 102 and the digital camera 101 via the communication means 103. 220 is an input unit such as a keyboard or mouse for operating the client terminal 102. 221 is a display unit such as a display for showing the input status of the client terminal 102.

[0037] 222 is a system bus for connecting 215 to 221 of each block that makes up the client terminal 102. 223 is the CPU for overall control of the learning server 104. 224 is the HDD that holds the programs and data necessary for the CPU 223 to operate. 225 is RAM for temporarily expanding the programs read by the CPU 223 from the HDD 224, executing the programs from there, and temporarily holding operational data.

[0038] The 226 is a GPU specifically designed for data processing, enabling high-speed image processing and matrix operations, and the processing of larger amounts of data.

[0039] The GPU226 can perform calculations more efficiently by processing data in parallel, so it is effective to use the GPU226 when performing training multiple times using learning models such as deep learning.

[0040] In the learning server 104, the learning process uses the GPU 226 in addition to the CPU 223. Specifically, when executing a learning program that includes a learning model, the CPU 223 and GPU 226 work together to perform calculations and carry out the learning process.

[0041] Furthermore, the learning process may be performed using only CPU223 or GPU226.

[0042] 227 is a NIC for communication with client terminals 102 and data server 105 via the local network 106.

[0043] 228 is an input unit such as a keyboard or mouse for operating the learning server 104. 229 is a display unit such as a display for showing the input status of the learning server 104. 230 is a system bus for connecting 223 to 229 of each block that makes up the learning server 104.

[0044] 231 is the CPU for overall control of the data server 105. 232 is the HDD that holds the programs and data necessary for the CPU 231 to operate. 233 is the RAM that temporarily expands the programs read by the CPU 231 from the HDD 232, executes the programs from there, and also temporarily holds operational data.

[0045] 234 is a NIC for communication with client terminals 102 and learning server 104 via the local network 106. 235 is an input unit such as a keyboard or mouse for operating the data server 105. 236 is a display unit such as a display for showing the input status of the data server 105.

[0046] 237 is a system bus for connecting 231 through 236 of each block that makes up the data server 105.

[0047] Using Figure 3, we will explain the software configuration that is realized by utilizing the hardware resources and programs shown in the hardware configuration diagram of Figure 2.

[0048] 301 is a camera control unit that controls the entire digital camera 101 by reading a program from ROM 202, which holds a program for controlling the digital camera 101, and deploying a part of the program to RAM 203.

[0049] The camera control unit 301 controls the entire digital camera based on input from the input unit 207. For example, it processes video from the rear camera 212 and the front camera 213 using the camera engine 209, and displays image data stored in the memory unit 210 on the display unit 208.

[0050] 302 is an estimation unit that uses a pre-trained model to estimate the input data using GPU204 or CPU201.

[0051] 303 is a data acquisition unit that acquires image data processed by the camera engine 209 from video footage from the rear camera 212 and the front camera 213.

[0052] The data acquisition unit 303 receives the acquired image data as input for the estimation unit 302 to perform estimation.

[0053] 304 is a data transmission and reception unit that transmits image files to the client terminal 102 via I / F 206 in order to transfer data to the data server 105, which stores the learning data for learning on the learning server 104.

[0054] The image file is learning data that includes image data processed by the camera engine 209 from video from the rear camera 212 and the front camera 213, and history data of user operations on the digital camera 101 via the input unit 207.

[0055] Furthermore, the data transmission / reception unit 304 also functions as a data transmission / reception unit that receives data from the client terminal 102 via the I / F 206 in order to acquire trained model data trained on the learning server 104.

[0056] 305 is a client terminal control unit that controls the entire client terminal 102 by reading a program from the HDD 216, which holds a program for the CPU 215 to control the client terminal 102, and by loading a part of the program into the RAM 217.

[0057] The client terminal control unit 305 receives training data from the digital camera 101 using the input unit 220 while the user is looking at the display unit 221, and instructs the data server 105 to send the image file.

[0058] Furthermore, the client terminal control unit 305 receives trained model data from the learning server 104 using the input unit 220 while the user is looking at the display unit 221, and instructs the digital camera 101 to transmit the trained model data.

[0059] 306 is a data transmission and reception unit that receives learning data transmitted from the digital camera 101 via I / F219 and transmits it to the data server 105 via NIC218.

[0060] Furthermore, the data transmission / reception unit 306 also functions as a data transmission / reception unit that receives trained model data learned by the learning server 104 via the NIC 218 and transmits it to the digital camera 101 via the I / F 219.

[0061] 307 is a data server control unit that controls the entire data server 105 by reading a program from the HDD 232, which holds a program for the CPU 231 to control the data server 105, and deploying a portion of the program to the RAM 233.

[0062] The data server control unit 307 stores the training data when it is sent from the client terminal 102, and controls the entire data server to send the training data when the training server 104 requests it to send training data.

[0063] 308 is a data collection / provision unit that receives learning data transmitted from client terminal 102 via NIC234 and collects the data.

[0064] 309 is a data storage unit for saving the learning data received by NIC234 to HDD232.

[0065] Furthermore, the data storage unit 309 reads data from the HDD 232 and passes it to the NIC 234 in order to provide learning data for the learning server 104 to learn from.

[0066] The data collection / provision unit 308 transmits learning data from NIC234 to the learning server 104 for the learning server 104 to learn from.

[0067] 310 is a learning server control unit that controls the entire learning server 104 by reading the program from the HDD224, which holds the program for the CPU223 to control the learning server 104, and deploying a part of the program to the RAM225.

[0068] The learning server control unit 310 receives learning data from the data server 105 using the input unit 228 while the user is looking at the display unit 229, and issues instructions for the learning process of the learning model.

[0069] Furthermore, the learning server control unit 310 instructs the digital camera 101 to send the trained model data to the client terminal 102.

[0070] 311 is a data transmission and reception unit that receives training data sent from data server 105 via NIC227.

[0071] Furthermore, the data transmission / reception unit 311 transmits the trained model data learned by the learning server 104 via the NIC 227.

[0072] 312 is a data management unit that allows the CPU 223 to determine and manage whether to use the training data received by the data transmission / reception unit 311 as training data, or to transmit the trained model data output from the data transmission / reception unit 311.

[0073] 313 is a training data generation unit that expands the received data into RAM225 as training data in the data management unit 312, and processes the data to determine which data to use as training data and which data to use as verification data.

[0074] The data generated by the learning data generation unit 313 is stored in RAM 225 or HDD 224.

[0075] 314 is a learning unit that uses the data generated by the learning data generation unit 313 and stored in the RAM 225 or HDD 224 to perform learning processing using the GPU 226 or CPU 223.

[0076] 315 is a data storage unit that stores the trained model, which has been trained by the learning unit 314, as a trained model in the HDD 224.

[0077] Figure 4 will be used to explain the learning model and the content it estimates.

[0078] 401 is a learning model that uses neural networks and the like.

[0079] 402 is live view image data captured by the rear camera 212 and front camera 213 of the digital camera 101, which are input to the learning model 401.

[0080] 403 is a user preference image determination, in which live view image data 402 is input into the learning model 401 to determine whether it is an image preferred by the user operating the digital camera 101.

[0081] Using Figure 5, we will explain the operation of a system to which the present invention can be applied, utilizing the structure of the learning model shown in Figure 4.

[0082] In step 501, training data is obtained, which is an image file created by associating image data captured by the user with the rear camera 212 or front camera 213 with operation history data performed by the user using the input unit 207 up to the point of capture.

[0083] In 502, the learning data acquired in 501 is transmitted from the digital camera 101 to the client terminal 102.

[0084] In step 503, the training data acquired in step 501 is sent from the client terminal 102 to the data server 105.

[0085] In step 504, the data server 105 stores the data for collection as training data.

[0086] In step 505, the client terminal 102 requests the learning server 104 to send trained model data to the digital camera 101.

[0087] In step 506, when the training server 104 receives a request for trained model data from the client terminal 102, it requests the data server 105 to send training data in order to train the model.

[0088] In step 507, the data server 105 sends the stored training data when it receives a request from the training server 104 to send training data.

[0089] In step 508, the learning server 104 uses the learning data received from the data server 105 to perform the learning process on the learning model.

[0090] In step 509, once the training process for the training model is complete, the training server 104 sends the data of the trained model to the client terminal 102.

[0091] In step 510, the client terminal 102 transmits the trained model data received from the training server 104 to the digital camera 101.

[0092] In 511, the digital camera 101 stores the trained model data received from the client terminal 102.

[0093] From step 501 to 511, the trained model described in Figure 4 is installed in the digital camera 101.

[0094] In 512, the user performs operations such as taking pictures using the digital camera 101.

[0095] When the user is operating the live view function on the digital camera 101, the learning model described in Figure 4 is used to estimate whether the live view image is to the user's liking. If it is estimated to be to the user's liking, the image is stored in the storage unit 210.

[0096] This allows the system to automatically capture and record images of the user's preferences while they are viewing review images, without the user having to consciously take a picture.

[0097] The operation flow of the digital camera 101 will be explained using Figure 6(a).

[0098] In S601, the digital camera 101 determines whether it is performing normal camera operation, playback operation, or storage operation of trained model data.

[0099] In S601, if it was a normal camera operation, the process proceeds to S602; if it was a playback operation, it proceeds to S615; and if it was a storage operation for trained model data, it proceeds to S613.

[0100] In S602, an operation buffer is allocated in RAM203 to temporarily store the user's operation history, the values ​​in the operation buffer are deleted, and the process transitions to S603.

[0101] In S603, the movement detection unit 205 is used to determine whether the user has changed the direction of the digital camera 101 beyond a predetermined value, or has moved it physically beyond a predetermined value. If the change in direction or movement is beyond the predetermined value, the process proceeds to S602; otherwise, it proceeds to S604.

[0102] In S604, it is determined whether the user has performed an operation using the input unit 207. If an operation has been performed, the process proceeds to S605; otherwise, the process proceeds to S603.

[0103] Examples of operations performed by the user using the input unit 207 include changing the camera's shutter speed and aperture value, changing the shooting mode, looking through the display unit 221, and focusing. Operations for zooming and focusing are also included.

[0104] In S605, the system determines whether the user has performed a switching or selection operation using the rear camera 212 or the front camera 213, or a playback operation from the input unit 207. If a switching or playback operation is performed, the system proceeds to S602.

[0105] In S605, the system determines whether the user is performing a switching operation using the rear camera 212 or the front camera 213, or a playback operation using the input unit 207. If a switching operation is performed, the system proceeds to S606; if a playback operation is performed, the system proceeds to S615.

[0106] In S606, it is determined whether the user performed a shooting operation using the release switch or other part of the input unit 207. If a shooting operation was performed, the process proceeds to S607; otherwise, it proceeds to S609.

[0107] In S607, the timer unit 214 retrieves time information of user operations, and the process proceeds to S608.

[0108] Examples of time information include the time interval between user actions, the time of the action, and the duration of the action.

[0109] In S608, the time information acquired in S607 and the type of operation performed by the user through the input unit 207 are stored in the operation buffer, and the process proceeds to S603.

[0110] In S609, the image data captured by the rear camera 212 or the front camera 213, along with the user's operation history data stored in the operation buffer up to the point of capture, are stored as a single image file in the storage unit 210 and then transferred to S610.

[0111] Examples of storage locations for operation history data include the Exif data area of ​​image files.

[0112] In S610, the system determines whether the shooting operation is continuous shooting. If it is continuous shooting, it proceeds to S609; otherwise, it proceeds to S611.

[0113] In S611, the image file is sent to client terminal 102 and then transferred to S612.

[0114] In S612, the system determines whether to continue camera operation. If it does, it proceeds to S602; otherwise, it terminates.

[0115] If S601 is an operation to store trained model data, the process proceeds to S613.

[0116] In S613, the system remains in S613 until it can receive data for the trained model from client terminal 102. Once the trained model data is obtained, it proceeds to S614.

[0117] In S614, the trained model data received in S613 is stored in GPU204 or RAM203, and the process terminates.

[0118] In S615, the digital camera 101 reads still images and video files from the memory unit 210, plays the still images and videos on the display unit 208, and then proceeds to S601.

[0119] The operation flow of client terminal 102 will be explained using Figure 6(b).

[0120] In S621, it is determined whether an image file has been received from the digital camera 101. If no image file has been received, the process proceeds to S622.

[0121] In S622, a decision is made as to whether to send the trained model data to the digital camera 101. If it is decided to send the trained model data, the process proceeds to S623.

[0122] In S623, a request is made to the training server 104 to send the trained model data, and the process proceeds to S624.

[0123] In S624, the system remains at S624 until the trained model data is sent from the training server 104 and can be received. Once reception is complete, the system proceeds to S625.

[0124] In step S625, the trained model data received in step S624 is transmitted to the digital camera 101, and the process ends.

[0125] In S621, it is determined whether an image file has been received from the digital camera 101. If an image file has been received, the process proceeds to S626.

[0126] In S626, the image file received in S621 is sent to the data server 105, and the process ends.

[0127] In S622, a decision is made as to whether to send the trained model data to the digital camera 101. If the trained model data is not to be sent, the process terminates.

[0128] Figure 6(c) will be used to explain the operation flow of the data server 105.

[0129] In S631, it is determined whether an image file has been received from the client terminal 102. If an image file has been received, the process proceeds to S632; otherwise, the process proceeds to S633.

[0130] In S632, image files received from client terminal 102 are stored in HDD232 and then transferred to S633.

[0131] In S633, the system checks whether the learning server 104 has requested the transmission of image files, which are training data. If there is a request to send training data, the system proceeds to S634. If there is no request to send training data, the system proceeds to S631.

[0132] In S634, the image files stored in HDD232 are read, and the image files are sent to the learning server 104 as training data, before proceeding to S631.

[0133] Figure 6(d) is used to explain the operation flow of the learning server 104.

[0134] In S641, the system checks whether the client terminal 102 has requested the transmission of trained model data. If there is no request for trained model data, the system remains in S641. If there is a request for trained model data, the system proceeds to S642.

[0135] In S642, a request is made to the data server 105 to send training data, and the process proceeds to S643.

[0136] In S643, the system remains in S643 until training data is received from data server 105, and then moves to S644 once the training data is received.

[0137] In S644, the operation history data of image files, which is training data acquired from data server 105, is retrieved, weighting is performed on each operation in the operation history to create training labels, and then the process moves to S645.

[0138] In S645, the image data from the image file and the labels created in S644 are input into the learning model, and the process moves to S646.

[0139] In S646, the training data input to the learning model is used to train it using GPU226 or CPU223, and then the process moves to S647.

[0140] In S647, all training data obtained from data server 105 is entered, and it is determined whether training is complete. If not all data has been entered, the process proceeds to S644; if all data has been entered, the process is considered complete and the process proceeds to S648.

[0141] In S648, the trained model data is sent to the client terminal 102 as trained model data, and the process ends.

[0142] Figure 7 illustrates the operation flow of the digital camera 101 that has stored the trained model data.

[0143] In S701, it is determined whether the power switch on the input unit 207 of the digital camera 101 is ON. If it is ON, the digital camera 101 is started and the process moves to S702; otherwise, the process remains in S701.

[0144] In S702, a live view operation is performed to display the captured data acquired by the rear camera 212 or the front camera 213 on the display unit 208, and then the process transitions to S703.

[0145] In S703, the data captured by the rear camera 212 or the front camera 213, acquired in S702, is input into the learning model 401, and the process moves to S704.

[0146] In S704, using GPU204 or CPU201, an estimation calculation is performed to determine if the input image is to the user's liking, and then the process proceeds to S705.

[0147] In S705, if the estimation calculation determines that the input image is to the user's liking, the process proceeds to S706; otherwise, the process proceeds to S702.

[0148] In S706, the system performs an image capture operation to store the input image in the memory unit 210, and then proceeds to S702.

[0149] Figure 9 illustrates the data structure of the operation history data section.

[0150] 901 is an operation sequence area that indicates the order in which the user performed operations.

[0151] 902 is the area representing the type of operation performed by the user.

[0152] For example, this includes operations such as changing the shutter speed, changing the aperture value, and changing the shooting mode.

[0153] 903 is the operation time domain, which represents the time information of the operations performed by the user.

[0154] Figure 10 illustrates the scoring table for the operations involved in creating learning labels.

[0155] 1001 indicates the type of operation.

[0156] 1002 indicates the score corresponding to operation type 1001.

[0157] For example, if the operation is to change the shutter speed, the score will be 2 points.

[0158] Figure 8 illustrates the flow of creating the training labels described in S644.

[0159] In S801, the operation type in the operation type area 902 and the operation time information in the operation time area 903 are obtained from the operation history data section of the image files, which are training data acquired from the data server 105, starting with the lowest order value shown in the operation order area 901. After that, the process moves to S802.

[0160] In S802, if the type of operation acquired is not a shooting operation, the process proceeds to S803.

[0161] In S803, the system retrieves points corresponding to the type of operation from the operation point table in Figure 10, performs scoring based on the operation, and then proceeds to S804.

[0162] In S804, weighting is performed based on operation time information, such as assigning a larger weight to operations with longer operation times, before proceeding to S805.

[0163] In S805, weighting is applied based on the order of operations, such as giving more weight to operations that are closer to the shooting operation, before transitioning to S801.

[0164] In S802, if the type of operation acquired is a shooting operation, the process proceeds to S806. In S806, all the points calculated up to that point are added together to create the learning label.

[0165] As a result, while the user is using the digital camera 101 in live view mode, the camera can automatically take a picture if the live image matches the user's preference, thus preventing the user from missing out on capturing their desired image.

[0166] (Other examples) The present invention can also be realized by supplying a program that implements one or more of the functions of the above-described embodiments to a system or device via a network or storage medium, and by having one or more processors in the computer of that system or device read and execute the program. It can also be realized by a circuit (e.g., an ASIC) that implements one or more functions.

Claims

1. A first imaging means for acquiring an image, A second imaging means for acquiring an image, An input section for the user to operate the device, Timer section, It includes a storage unit that stores the captured image and the operation history of the input unit in association with each other. An imaging device characterized in that, when a user performs a selection operation from the input unit to the imaging device to switch between using the first imaging means and using the second imaging means, the operation history prior to the selection operation is deleted and stored in the storage unit.

2. The imaging device according to claim 1, characterized in that the timer unit is used to store time information when the input unit is operated in the operation history.

3. The imaging device according to claim 1, characterized in that the storage area of ​​the storage unit that stores the aforementioned operation history in association is an Exif area.

4. The imaging device according to claim 2, characterized in that the aforementioned time information is the time interval between each operation.

5. The system includes an estimation unit that estimates whether the image data from either the first imaging means or the second imaging means is to the user's liking. The imaging device according to any one of claims 1 to 4, characterized in that it takes a photograph if the estimation unit estimates that it is the user's preference.

6. The imaging device according to claim 5, characterized in that the estimation unit creates preference labels from the operation history associated with the stored images, and the trained model is trained using the images and the created labels.

7. The imaging device according to claim 6, characterized in that when creating a preferred label from the operation history, the preferred weighting value is changed based on at least one of the type of operation in the operation history, the time information of the operation history, or the time interval of the operation history.

8. A control method for an imaging device having a first imaging means for acquiring an image, a second imaging means for acquiring an image, and an input unit for a user to operate the device, The system includes a storage step that associates the captured image with the operation history of the input unit, A control method for an imaging device, characterized in that, when a user performs a selection operation from the input unit to the imaging device to switch between using the first imaging means and the second imaging means, the operation history prior to the selection operation is deleted and stored in the storage unit.

9. A computer-readable program for causing a computer to function as one of the means of an imaging device according to any one of claims 1 to 7.

Images