A gesture shooting method and device based on multi-person scene

Abstract

The present invention discloses a gesture shooting method and device based on a multi-person scene, the method comprising: obtaining a preview shooting picture containing multiple persons captured by a camera; determining a detection area for triggering a shooting operation in the preview shooting picture; wherein the area of ​​the detection area is smaller than the preview shooting picture; detecting the gestures of the person in the detection area, and controlling the camera to shoot when a preset shooting gesture is detected. The present invention can more accurately identify the shooting gesture through a designated detection area, so that the shooting operation can be completed without detecting the shooting gesture of the entire picture, avoiding the situation where the shooting is interfered with by multiple gestures of multiple persons, improving the accuracy of shooting in multi-person shooting scenes, and better meeting the needs of users.

Description

Technical Field

[0001] The present invention relates to the field of gesture shooting technology, and in particular to a gesture shooting method and device based on a multi-person scene. Background Art

[0002] Currently, the shooting equipment and shooting methods on the market basically support users to achieve shooting through gestures, that is, users do not need to touch the screen with their fingers, but automatically complete the shooting when it detects that there are people in the picture or people make shooting gestures.

[0003] Existing gesture recognition technology is mainly aimed at single-person scenarios, where a single user performs a gesture to trigger the corresponding operation. However, in multi-person scenarios, due to the presence of multiple users, different people may unconsciously make multiple gestures, resulting in the incorrect triggering of corresponding shooting operations. That is, the existing shooting terminal cannot select a unique and correct shooting gesture to respond to trigger shooting, resulting in operational confusion or operational errors when multiple people are shooting, which cannot meet user needs. Summary of the invention

[0004] The embodiments of the present invention provide a gesture shooting method and device based on a multi-person scenario, which can effectively solve the problem in the prior art that a unique and correct shooting gesture action cannot be selected to respond to trigger shooting, resulting in operation confusion or operation errors when multiple people are shooting, and cannot meet user needs.

[0005] An embodiment of the present invention provides a gesture shooting method based on a multi-person scene, comprising:

[0006] Get the preview image of multiple people captured by the camera;

[0007] Determining a detection area for triggering a shooting operation in the preview shooting picture; wherein the area of ​​the detection area is smaller than the preview shooting picture;

[0008] The gesture of the person is detected in the detection area, and when a preset shooting gesture is detected, the camera is controlled to shoot.

[0009] Preferably, determining a detection area for triggering a shooting operation in the preview shooting picture includes:

[0010] A detection area for triggering a shooting operation is generated at a preset position in the preview shooting picture according to a preset size.

[0011] Preferably, determining a detection area for triggering a shooting operation in the preview shooting picture includes:

[0012] Calculate the distance between each character and the center of the preview shot image, and mark the character with the shortest distance as the main control character;

[0013] A detection area for triggering a shooting operation is generated according to the area occupied by the main control character in the preview shooting picture.

[0014] Preferably, determining a detection area for triggering a shooting operation in the preview shooting picture includes:

[0015] When a gesture of a person is detected as a first preset master gesture within a first preset gesture reaction time, the person is marked as a master person;

[0016] A detection area for triggering a shooting operation is generated according to the area occupied by the main control character in the preview shooting picture.

[0017] Preferably, before detecting that a person's gesture is the preset master gesture within the first preset gesture reaction time, the method further includes:

[0018] Generate first prompt information for prompting a user to make a first preset master control gesture, so that the user makes the first preset master control gesture according to the first prompt information; wherein the first prompt information includes: a first preset gesture reaction time and a graphic of the first preset master control gesture.

[0019] Preferably, it also includes:

[0020] If the first preset master control gesture is not detected within the first preset gesture reaction time, generating second prompt information for prompting the user to make a second preset master control gesture;

[0021] Among them, the second prompt information includes: a graphic of a second preset master control gesture and a second preset gesture reaction time; the first preset master control gesture is different from the second preset master control gesture, and the second preset gesture reaction time is longer than the first preset gesture reaction time.

[0022] Preferably, after controlling the camera to shoot, the method further comprises:

[0023] The picture currently captured by the camera is used as a pre-saved picture, and prompt information of a preset saving gesture is displayed in the pre-saved picture, so that the user can make a corresponding gesture according to the prompt information of the preset saving gesture;

[0024] When a preset saving gesture is detected in the pre-saved screen, the pre-saved screen is saved as a picture or a video.

[0025] Preferably, it also includes:

[0026] Inputting the selected picture into a preset background replacement model so that the background replacement model performs semantic estimation on the characters in the selected picture to obtain positions of several characters; wherein the selected picture is a preview shot picture or a pre-saved picture;

[0027] Extracting the character features corresponding to each character, and segmenting the character and the background other than the character based on the position of each character and the character features corresponding to each character, to obtain a number of segmented characters;

[0028] The preset background is used as the background of the segmented person, and the updated selected picture is synthesized according to the segmented person and the preset background.

[0029] An embodiment of the present invention provides a gesture shooting method based on a multi-person scene, comprising:

[0030] Acquire a first preview shot containing multiple persons captured by a first camera;

[0031] Determining a first detection area for triggering a shooting operation in the first preview shooting picture; wherein the area of ​​the first detection area is smaller than the first preview shooting picture;

[0032] The gesture of the person is detected in the first detection area, and when a preset shooting gesture is detected, the second camera is controlled to shoot.

[0033] Based on the above method embodiments, the present invention provides corresponding device embodiments.

[0034] An embodiment of the present invention provides a gesture shooting device based on a multi-person scene, comprising: a picture acquisition module, a detection area generation module and a shooting control module;

[0035] The picture acquisition module is used to acquire a preview picture containing multiple characters captured by the camera;

[0036] The detection area generation module is used to determine a detection area for triggering a shooting operation in the preview shooting picture; wherein the area of ​​the detection area is smaller than the preview shooting picture;

[0037] The shooting control module is used to detect the gestures of the person in the detection area, and control the camera to shoot when a preset shooting gesture is detected.

[0038] The following beneficial effects are achieved by implementing the present invention:

[0039] The embodiment of the present invention provides a gesture shooting method and device based on a multi-person scene, the method comprising: obtaining a preview shooting picture containing multiple people captured by a camera; determining a detection area for triggering a shooting operation in the preview shooting picture; wherein the area of ​​the detection area is smaller than the preview shooting picture; detecting the gesture of the person in the detection area, and controlling the camera to shoot when a preset shooting gesture is detected. Compared with the prior art, after obtaining the preview shooting picture containing multiple people captured by the camera, the present invention can determine a detection area for triggering a shooting operation in the preview shooting picture, so as to more accurately identify the shooting gesture through the designated detection area. Because the area of ​​the detection area is smaller than the preview shooting picture, the present invention can achieve the shooting operation without detecting the shooting gesture of the entire picture, avoiding the situation where the shooting is interfered by the multiple gestures of multiple people, improving the accuracy of shooting in the multi-person shooting scene, and better meeting the needs of users. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flowchart of a gesture shooting method based on a multi-person scene provided by an embodiment of the present invention.

[0041] Figure 2 This is another flowchart of a gesture-based shooting method provided by another embodiment of the present invention.

[0042] Figure 3 It is a structural schematic diagram of a gesture shooting device based on a multi-person scene provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] like Figure 1 FIG. 1 is a flow chart of a gesture shooting method based on a multi-person scene provided by an embodiment of the present invention. The gesture shooting method based on a multi-person scene includes:

[0045] Step S1: obtaining a preview shot image containing multiple persons captured by a camera;

[0046] Step S2: determining a detection area for triggering a shooting operation in the preview shooting picture; wherein the area of ​​the detection area is smaller than the preview shooting picture;

[0047] Step S3: Detecting the gesture of the person in the detection area, and controlling the camera to shoot when a preset shooting gesture is detected.

[0048] Regarding step S1, in a preferred embodiment, after the camera is turned on, a preview of the shooting picture can be displayed to the user so that the user can make a corresponding shooting gesture.

[0049] For step S2, in a preferred embodiment, a detection area for triggering a shooting operation is determined in the preview shooting screen, and the area of ​​the detection area is smaller than the entire preview shooting screen, so that a designated detection area can be displayed in the preview shooting screen to more accurately identify the shooting gesture.

[0050] In a preferred embodiment, if there is only a single character in the preview interface, then the character is the main control, and the shooting can be controlled by gestures. If there are multiple characters in the preview interface, it is necessary to determine one of the characters as the main control character.

[0051] The method for selecting the main control character can be determined based on the number of people in the picture or the center of the picture, for example, the person in the middle of the crowd or closest to the center of the picture can be selected as the main control; or a special gesture can be displayed in the preview interface, and the character who responds to the gesture within a specified time can be observed to determine that the character is the main control.

[0052] Specifically, the present invention can, when determining a detection area for triggering a shooting operation, generate a detection area for a main control character or a gesture detection area for generating a gesture collection frame; wherein, after determining the main control character, based on a gesture detection algorithm, detect whether the main control character has made a correct shooting gesture, and if so, trigger shooting; when a shooting gesture appears in the gesture detection area, it is determined as an input trigger instruction, and the shooting operation is triggered.

[0053] It should be noted that the present invention can train a gesture detection model according to the set gestures. The model can be a single-target detection model or a multi-target detection model, which is used to recognize and detect different gestures, such as the OK gesture, the thumbs-up gesture, the yes gesture, and the six gestures.

[0054] For step S3, in a preferred embodiment, when a preset shooting gesture is detected in the detection area, a shooting instruction is triggered to control the camera to shoot.

[0055] The shooting triggering conditions of the present invention: In order to avoid false triggering, the shooting triggering conditions can be set. One way is to determine the main control character and then detect whether the main control has made a control gesture based on the gesture detection algorithm, and trigger shooting if so; another way is to not preset the main control character, but to give a gesture collection frame in the picture, and only when the shooting gesture appears in the frame, it is determined as an input trigger instruction, thereby triggering shooting.

[0056] Because the area of ​​the detection area is smaller than the preview shooting screen, the present invention can complete the shooting operation without detecting the shooting gesture of the entire screen, avoiding the interference of multiple gestures of multiple people on the shooting, improving the accuracy of shooting in multi-person shooting scenes, and better meeting the needs of users.

[0057] In a preferred embodiment, determining a detection area for triggering a shooting operation in the preview shooting picture includes:

[0058] A detection area for triggering a shooting operation is generated at a preset position in the preview shooting picture according to a preset size.

[0059] The present invention can preset a gesture detection area in the preview interface and highlight the area; when a preset shooting gesture appears in the gesture detection area, it is determined as an input trigger instruction and triggers the shooting operation. The specific steps are as follows:

[0060] a) activating the first camera to collect images, and providing a detection area for triggering a shooting operation at a certain position in the preview interface of the display device;

[0061] b) informing the user through voice or text prompts that the shooting can be activated when a trigger gesture is collected in the detection area;

[0062] In a preferred embodiment, determining a detection area for triggering a shooting operation in the preview shooting picture includes:

[0063] Calculate the distance between each character and the center of the preview shot image, and mark the character with the shortest distance as the main control character;

[0064] A detection area for triggering a shooting operation is generated according to the area occupied by the main control character in the preview shooting picture.

[0065] The present invention can count the number of people in real time based on the human body detection results in the picture, and display the number of people in the preview interface. If the number of people is 1, the person is assumed to be the main controller, and the person performs the gesture photo operation; if the number of people is greater than 1, it is necessary to further determine the main controller;

[0066] The main control character selection method of the present invention: The main control character can be determined by the following method:

[0067] a) According to the number of people and their distribution in the picture obtained by the human body detection algorithm, determine the person corresponding to the center of the crowd as the main control. For example, if there are 7 people standing in a row, the 4th person may be the center of the crowd;

[0068] b) According to the distance between the human body and the center of the picture, the person closest to the center of the picture is selected as the main control;

[0069] That is, the main controller is set to the person who is closer to the center of the crowd and the closest to the center of the screen. For example, if there are 8 people in a queue, the 4th and 5th person who is closer to the center of the screen will be the main controller;

[0070] After the main person is determined, the logo window of the person can be set to a more eye-catching color to prompt the user so that the user can know who the main person is. At the same time, the detection area for triggering the shooting operation is generated according to the area or outline occupied by the main person in the preview shooting screen, and then the detection area of ​​the main person can be continuously detected to see if a preset shooting gesture appears. If a shooting gesture is detected, the shooting instruction is triggered, and the screen displayed in the preview interface is saved as a picture or video.

[0071] Through the above steps, it is possible to control shooting through gestures in multi-person scenes. Giving the main control to people in important positions can make shooting operations more reasonable and avoid false triggering.

[0072] The method for selecting the master controller may also include specific gesture response: displaying a specific gesture on the preview interface, and detecting in real time who in the camera image responds to the gesture within a specified time, and then determining that person as the master controller.

[0073] In a preferred embodiment, determining a detection area for triggering a shooting operation in the preview shooting picture includes:

[0074] Generate first prompt information for prompting a user to make a first preset master control gesture, so that the user makes the first preset master control gesture according to the first prompt information; wherein the first prompt information includes: a first preset gesture reaction time and a graphic of the first preset master control gesture;

[0075] When a gesture of a person is detected as a first preset master gesture within a first preset gesture reaction time, the person is marked as a master person;

[0076] A detection area for triggering a shooting operation is generated according to the area occupied by the main control character in the preview shooting picture.

[0077] If a person's gesture is not detected as the preset main control gesture within the first preset gesture reaction time, a third prompt message is generated to prompt the user to make the preset main control gesture; wherein the third prompt message includes the second preset gesture reaction time and the second preset main control gesture, the first preset main control gesture is different from the second preset main control gesture, and the second preset gesture reaction time is longer than the first preset gesture reaction time.

[0078] Through the above method, the main control person can be determined through specific gesture response or gesture detection area trigger, and the gesture shooting function can be realized. Such a design can increase the flexibility and interactivity of the shooting operation, allowing users to control the shooting process more conveniently.

[0079] The present invention can determine whether someone has made a master control gesture by setting a preset gesture reaction time. When no master control gesture is detected within the preset gesture reaction time, a different master control gesture and time can be generated, such as a simpler or easier to recognize master control gesture, and the gesture reaction time of the second recognition is increased to help users respond better and improve the accuracy of recognition.

[0080] The specific steps are as follows:

[0081] a) Through voice or text prompts, the main control character is required to make specific gestures within a specified time and activate the first camera for collection;

[0082] b) Detect whether there is gesture input within the specified time. If there is gesture input, mark the person, calibrate and track him;

[0083] c) The electronic device notifies the user through voice or screen text that the main person has been selected, and highlights the main person's body mark window. At the same time, the main person is required to make a shooting gesture within a specified time to trigger the shooting;

[0084] d) Continue to track whether the main control person makes a shooting gesture. If it is detected that the main control person makes a shooting gesture, activate the camera to shoot; if no shooting is performed after a specified time, you can re-enter the main control selection stage.

[0085] Through the above method, the main control person can be determined through specific gesture response, and the gesture shooting function can be realized. Such a design can increase the flexibility and interactivity of the shooting operation, allowing the user to control the shooting process more conveniently.

[0086] The present invention can determine the main control person through the trigger of a specific number of people, position or gesture detection area. After determining the main control hand, it ensures that only the main control gesture is recognized as the shooting trigger instruction, avoiding the problem of false triggering, enhancing the controllability and reliability of shooting, and realizing accurate and stable gesture shooting function in multi-person scenes.

[0087] This application provides three shooting methods in multi-person scenes, a location-based method, a gesture response-based method, and a fixed gesture detection area-based method, so that users can choose according to the scene usage requirements to meet the needs of users. The technical solution of the present invention adopts a high-precision gesture recognition algorithm and a reliable action recognition model, which can accurately analyze and judge the user's gesture actions. By optimizing the algorithm and model, the stability of gesture recognition can be improved and the possibility of false triggering can be reduced.

[0088] An embodiment of the present invention further includes:

[0089] Inputting the selected picture into a preset background replacement model so that the background replacement model performs semantic estimation on the characters in the picture to obtain the positions of several characters; wherein the selected picture is a preview shot picture or a pre-saved picture, the background replacement process can be performed before or after shooting;

[0090] Extracting the character features corresponding to each character, and segmenting the character and the background other than the character based on the position of each character and the character features corresponding to each character, to obtain a number of segmented characters;

[0091] The preset background is used as the background of the segmented characters, and the updated selected picture is synthesized according to the segmented characters and the preset background; wherein the preset background is a pre-stored picture, or a background picture selected by the user.

[0092] The preview screen of the present invention can be processed by background replacement. The process includes the following steps:

[0093] The image from the first camera is fed into the background replacement model as input. The background replacement model uses an algorithm to automatically identify the human body and the background, cuts the human body out of the original background, and replaces the background. The replacement background can be a picture pre-stored in the system, and the user can select it through the system. The replaced image is displayed in the preview interface for the user to view.

[0094] The background replacement cutout model uses a neural network-based algorithm, which does not require foreground and background annotation and can achieve real-time cutout. The model is based on the following principles:

[0095] The network is divided into two branches: Branch 1 is used to perform semantic estimation on the image, locate the position of the person, and obtain a rough segmentation result of the person; Branch 2 extracts detailed information of the person through a shallow neural network.

[0096] The results of branch 1 and branch 2 are superimposed to obtain a cutout mask Ap. Based on the cutout mask Ap, the image is cutout to separate the person from the background, and the model can realize real-time cutout and real-time display.

[0097] The background replacement method is as follows, assuming that the input image is Pi, the output image is Po, the background is Pb, and Ap is the cutout mask output by the network: Po = Ap*Pi+(1-Ap)*Pb;

[0098] Background replacement can be performed before or after gesture shooting. If background replacement occurs before shooting, users can select the background through the preview interface and replace it in real time in the cutout model. If background replacement occurs after shooting, users can input the saved image into the background replacement model, perform cutout and background replacement, and then display it in the preview interface, where users can select the replaced background. Through this background replacement process, users can view images with replaced backgrounds in the preview interface, which increases the personalization and creativity of the shooting effect.

[0099] In a preferred embodiment, after controlling the camera to shoot, the method further includes:

[0100] The picture currently captured by the camera is used as a pre-saved picture, and prompt information of a preset saving gesture is displayed in the pre-saved picture, so that the user can make a corresponding gesture according to the prompt information of the preset saving gesture;

[0101] When a preset saving gesture is detected in the pre-saved screen, the pre-saved screen is saved as a picture or a video.

[0102] After shooting is completed, the user can preview the shooting results in the preview interface to confirm the shooting effect. If the user is satisfied with the shooting results, the system will prompt the user to save. In order to trigger the saving operation, a preset gesture can be given, such as a "confirm saving" gesture. The user can perform the corresponding gesture to confirm and save the shooting results.

[0103] The present invention can also realize the functions of downloading, QR code sharing and adding watermarks:

[0104] 1. After saving, the system will generate a download link or QR code. Users can obtain the captured pictures or video files by scanning the QR code or clicking the download link. The download link can provide a one-time download, or generate a download link containing multiple shooting results based on the number of shots specified by the user.

[0105] 2. In the preview interface, users can choose to add a watermark to the shooting results. The watermark can be customized by the device or edited by the user. Users can enter any text sentence in the text box and select the specific location of the watermark. After entering the text and selecting the location, the system will display the watermark in the image of the preview interface in real time and preview the effect. Watermarks can include time watermarks, location watermarks, application watermarks, etc. to increase the information and personalized effect of the shooting results. In this way, users can preview the results after shooting and choose to save or add watermarks according to their satisfaction. At the same time, a download link or QR code sharing is provided to facilitate users to obtain the captured pictures or video files. Such functions enhance the user's ability to control and customize the shooting results.

[0106] An embodiment of the present invention provides a gesture-based shooting method, such as Figure 2 As shown, the gesture-based shooting method includes:

[0107] Acquire a first preview shot containing multiple persons captured by a first camera;

[0108] Determining a first detection area for triggering a shooting operation in the first preview shooting picture; wherein the area of ​​the first detection area is smaller than the first preview shooting picture;

[0109] The gesture of the person is detected in the first detection area, and when a preset shooting gesture is detected, the second camera is controlled to shoot.

[0110] The present invention relates to the collaborative use of multiple cameras, wherein the first camera is used for gesture recognition and master control selection, and the second camera is used for actual shooting. The first and second cameras can be swapped, and the user can select a suitable camera for shooting to obtain photos or videos at different angles and perspectives. This configuration increases the flexibility and selectivity of shooting, and improves the user's shooting experience and results.

[0111] It should be noted that the second camera can be separated from the electronic device and has a certain angle with the first camera. The position and angle of the second camera can be adjusted according to the scene, and this adjustment can be manual or automatic. The electronic device can turn on the first camera and the second camera at the same time and display their images in the preview interface.

[0112] The preview interface can choose to display only the image of the first camera, or display the images of the first camera and the second camera at the same time, and determine the main controller based on the image of the first camera.

[0113] The first and second cameras can be used as gesture detection inputs and can also be used for actual shooting. Users can set gestures by themselves, such as yes, like, six, etc.

[0114] When the first camera detects that the user performs a specific gesture, it can trigger the operation of controlling the second camera to shoot. Once the gesture is detected, the system can control the second camera to shoot and save the image of the second camera as a picture or video file. The functions of the first and second cameras can be swapped.

[0115] The second camera can be set as a gesture detection input to determine the position of the gesture, and automatically adjust the position and angle of the second camera to keep the gesture in the center of the picture. The second camera can be a SLR camera, an industrial camera, or a camera with different optical viewing angles, such as an ultra-telephoto camera, a telephoto camera, a normal camera, a wide-angle camera, and an ultra-wide-angle camera.

[0116] The gesture shooting method based on multi-person scenes of the present invention is applicable to gesture shooting in multi-person scenes. By analyzing and identifying the gestures of multiple users, the master hand can be determined, and the gestures of other people will not mistakenly trigger the shooting command, thus avoiding the problem of out-of-control of the scene.

[0117] Unlike the gesture recognition based on depth camera in the prior art, the technical solution of the present invention can use ordinary optical cameras for gesture recognition without relying on depth camera. This greatly expands the application scope of gesture photography and improves the flexibility of usage scenarios. In addition to the basic photo function, the technical solution of the present invention also supports multiple function extensions such as background replacement, beauty, adding watermarks, self-service download, etc. Users can trigger these functions through gestures, providing more shooting and post-processing options, and increasing user creativity and interactivity.

[0118] The present invention also introduces a second camera that can work in conjunction with the first camera. By opening two cameras at the same time and previewing, the user can select the appropriate camera to shoot and obtain photos or videos from different angles and perspectives. This configuration increases the flexibility and selectivity of shooting, and improves the user's shooting experience and results. Therefore, the technology of the present invention has the advantages of ordinary optical camera application, multi-functional expansion, stability and accuracy, and provides users with a more convenient, flexible and creative photo experience.

[0119] like Figure 3 As shown, based on the above-mentioned various embodiments of the gesture shooting method based on multi-person scenes, the present invention provides a corresponding device embodiment;

[0120] An embodiment of the present invention provides a gesture shooting device based on a multi-person scene, comprising: a picture acquisition module, a detection area generation module and a shooting control module;

[0121] The picture acquisition module is used to acquire a preview picture containing multiple characters captured by the camera;

[0122] The detection area generation module is used to determine a detection area for triggering a shooting operation in the preview shooting picture; wherein the area of ​​the detection area is smaller than the preview shooting picture;

[0123] The shooting control module is used to detect the gestures of the person in the detection area, and control the camera to shoot when a preset shooting gesture is detected.

[0124] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without creative work.

[0125] Those skilled in the art can clearly understand that, for the sake of convenience and simplicity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0126] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A gesture shooting method based on a multi-person scene, characterized in that: include: Get the preview image of multiple people captured by the camera; Generate first prompt information for prompting a user to make a first preset master control gesture within a first preset gesture reaction time, so that the user makes the first preset master control gesture according to the first prompt information; wherein the first prompt information includes: the first preset gesture reaction time and a graphic of the first preset master control gesture; If the first preset master control gesture is not detected within the first preset gesture reaction time, a second prompt message is generated to prompt the user to make a second preset master control gesture; wherein the second prompt message includes: a graphic of the second preset master control gesture and a second preset gesture reaction time; the first preset master control gesture is different from the second preset master control gesture, and the second preset gesture reaction time is longer than the first preset gesture reaction time; When detecting that a gesture of a person is a preset master gesture, marking the person as a master person; Generate a detection area for triggering a shooting operation according to the area occupied by the main control character in the preview shooting screen; wherein the area of ​​the detection area is smaller than the preview shooting screen; The gesture of the person is detected in the detection area, and when a preset shooting gesture is detected, the camera is controlled to shoot.

2. The gesture shooting method based on a multi-person scene as claimed in claim 1, characterized in that: After controlling the camera to shoot, the method further includes: The picture currently captured by the camera is used as a pre-saved picture, and prompt information of a preset saving gesture is displayed in the pre-saved picture, so that the user can make a corresponding gesture according to the prompt information of the preset saving gesture; When a preset saving gesture is detected in the pre-saved screen, the pre-saved screen is saved as a picture or a video.

3. The gesture shooting method based on a multi-person scene as claimed in claim 2, characterized in that: Also includes: Inputting the selected picture into a preset background replacement model so that the background replacement model performs semantic estimation on the characters in the selected picture to obtain positions of several characters; wherein the selected picture is a preview shot picture or a pre-saved picture; Extracting the character features corresponding to each character, and segmenting the character and the background other than the character based on the position of each character and the character features corresponding to each character, to obtain a number of segmented characters; The preset background is used as the background of the segmented person, and the updated selected picture is synthesized according to the segmented person and the preset background.

4. A gesture shooting method based on a multi-person scene, characterized in that: include: Obtaining a preview shot image containing multiple persons captured by the first camera; Generate first prompt information for prompting a user to make a first preset master control gesture within a first preset gesture reaction time, so that the user makes the first preset master control gesture according to the first prompt information; wherein the first prompt information includes: the first preset gesture reaction time and a graphic of the first preset master control gesture; If the first preset master control gesture is not detected within the first preset gesture reaction time, a second prompt message is generated to prompt the user to make a second preset master control gesture; wherein the second prompt message includes: a graphic of the second preset master control gesture and a second preset gesture reaction time; the first preset master control gesture is different from the second preset master control gesture, and the second preset gesture reaction time is longer than the first preset gesture reaction time; When detecting that a gesture of a person is a preset master gesture, marking the person as a master person; Generate a detection area for triggering a shooting operation according to the area occupied by the main control character in the preview shooting picture; wherein the area of ​​the detection area is smaller than the preview shooting picture; The gesture of the person is detected in the detection area, and when a preset shooting gesture is detected, the second camera is controlled to shoot.

5. A gesture shooting device based on a multi-person scene, characterized in that: include: Picture acquisition module, detection area generation module and shooting control module; The picture acquisition module is used to acquire a preview picture containing multiple characters captured by the camera; The detection area generation module is used to generate first prompt information for prompting a user to make a first preset master control gesture within a first preset gesture reaction time, so that the user can make the first preset master control gesture according to the first prompt information; wherein the first prompt information includes: the first preset gesture reaction time and a graphic of the first preset master control gesture; If the first preset master control gesture is not detected within the first preset gesture reaction time, a second prompt message is generated to prompt the user to make a second preset master control gesture; wherein the second prompt message includes: a graphic of the second preset master control gesture and a second preset gesture reaction time; the first preset master control gesture is different from the second preset master control gesture, and the second preset gesture reaction time is longer than the first preset gesture reaction time; When detecting that a gesture of a person is a preset master gesture, marking the person as a master person; Generate a detection area for triggering a shooting operation according to the area occupied by the main control character in the preview shooting screen; wherein the area of ​​the detection area is smaller than the preview shooting screen; The shooting control module is used to detect the gestures of the person in the detection area, and control the camera to shoot when a preset shooting gesture is detected.

Images