Video head fast shooting and analysis method based on AI algorithm

Abstract

The invention discloses a video head fast shooting and analysis method, system and device based on an AI algorithm and a medium. According to the technical scheme provided by the invention, external image information is acquired as an initial image according to a camera; analyzing the initial image, determining at least one image area contained in the initial image, and determining focusing information corresponding to each image area; determining a target image area from the at least one image area, and obtaining target focusing information corresponding to the target image area; and completing shooting based on the target focusing information. According to the invention, the AI algorithm is utilized, the currently acquired image is analyzed based on the historical image data, the focusing target of a photographer is intelligently judged, and the photographing focus of the image is determined, so that focusing and photographing are quickly completed.

Description

Technical Field

[0001] The present invention relates to the field of image capture, and in particular to a video camera snapshot capture and analysis method, system, computing device and computer storage medium based on an AI algorithm. Background Art

[0002] With the rapid development of artificial intelligence technology, the functions it can achieve are becoming more and more complex, and its application scenarios have gradually covered various fields. In the field of image shooting, it is also possible to intelligently segment images and extract specific parts based on deep learning technology.

[0003] Currently, the most common method is to perform face recognition in the viewfinder based on artificial intelligence during the shooting process, and to focus on the clarity of the face in the image to complete the photo. However, whether it is professional photographers, photography enthusiasts or ordinary people, the content they want to shoot is not limited to portraits. Therefore, when a large number of shooting targets are not portraits, this focusing method becomes an obstacle to the shooting process, prolonging the shooting time, and thus affecting the user experience. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a video camera quick shooting and analysis method based on AI algorithm and a corresponding video camera quick shooting and analysis system, computing device and computer storage medium based on AI algorithm.

[0005] According to one aspect of the present invention, a video recording and analysis method based on an AI algorithm is provided, the method comprising:

[0006] Acquire external image information according to the camera as an initial image;

[0007] Analyze the initial image, determine at least one image area included in the initial image, and determine focus information corresponding to each image area;

[0008] Determine a target image area from at least one image area, and obtain target focus information corresponding to the target image area;

[0009] Shooting is completed based on the target focus information.

[0010] In the above solution, the step of obtaining external image information as the initial image based on the camera further includes:

[0011] Receiving external image information based on a camera;

[0012] The received external image information is stored and whether it meets the preset judgment conditions is determined;

[0013] If yes, the external image information is set as the initial image; if no, the camera is reused to obtain new external image information.

[0014] In the above solution, the analysis of the initial image, determining at least one image area contained in the initial image, and determining the focus information corresponding to each image area further includes:

[0015] Construct image segmentation model based on neural network;

[0016] Using past historical images to train the image segmentation model, and obtaining a trained image segmentation model;

[0017] Analyzing the initial image based on the trained image segmentation model, and dividing the initial image into at least one image region;

[0018] For each image area, the corresponding focus information is determined.

[0019] In the above solution, determining the target image area from at least one image area and acquiring the target focus information corresponding to the target image area further includes:

[0020] Analyze and obtain image features of the initial image;

[0021] Based on the image features of the initial image, a comparison image similar to the initial image is selected from the historical images;

[0022] According to the contrast image, determining an image area corresponding to a focus in the contrast image;

[0023] By analyzing the selection of the image area corresponding to the focus in the comparison image, the target image area in the initial image is determined, and the target focus information corresponding to the target image area is obtained.

[0024] In the above solution, completing the shooting based on the target focus information further includes:

[0025] Get target focus information;

[0026] Adjusting the focus and focal length of the initial image according to the target focus information to generate a focused image;

[0027] In response to the user's focus determination operation completing the photographing, the focused image is determined as a final photographed image, and is output and stored.

[0028] In the above solution, determining the target image area from at least one image area and acquiring the target focus information corresponding to the target image area further includes:

[0029] According to the image features of the initial image, a plurality of first images having the same image features are determined from the historical images based on the image segmentation model;

[0030] Calculating the similarity between each first image and the initial image using the image segmentation model;

[0031] Determine the first image exceeding a preset similarity threshold as a comparison image;

[0032] According to the image areas corresponding to the focal points in each comparison image, statistics are performed to determine the number of images in which the focal points in the comparison image are located in different image areas;

[0033] Based on the number of images, the priority of the image area corresponding to the focus is determined, and the target image area in the initial image is determined based on the image area with the highest priority.

[0034] In the above scheme, the method further comprises:

[0035] The image partition model is based on the SAM model, and the weight matrix in the image partition model is adjusted based on the low-rank matrix, thereby adjusting the Query parameter Q′ and the Value parameter V′.

[0036] Q′=X in (M Q +ΔM Q )=X in (M Q +A Q B Q )

[0037] V′=X in (M V +ΔM V )=X in (M V +A V B V )

[0038] Among them, X in is the input matrix; M Q is the query weight matrix; ΔM Q A is the query weight adjustment matrix; Q , B Q is the query low-rank matrix; M V is the numerical weight matrix; ΔM V A is the numerical weight adjustment matrix; V , B V is a numerical low-rank matrix;

[0039] Based on the adjusted weight matrix, an adjusted attention weight is obtained, and then an adjusted image segmentation model is obtained, and based on the image features of the initial image, a plurality of first images having the same image features are determined from the historical images;

[0040] For the plurality of first images, based on the image features of the initial image and the image features of the first image, the similarity between the two is calculated,

[0041]

[0042] Among them, S(t begin ,t sample ) is the similarity value; t begin is the image feature of the initial image; t sample is the image feature of the first image; d represents the dimension of the image feature; c represents the sequence number of the current dimension; is the component of the image feature of the initial image in the current dimension; is the component of the first image feature in the current dimension.

[0043] According to another aspect of the present invention, a video camera snapshot and analysis system based on an AI algorithm is provided, comprising: an image acquisition module, a feature analysis module, a focus determination module and a shooting module; wherein:

[0044] The image acquisition module is used to acquire external image information as an initial image through a camera;

[0045] The feature analysis module is used to analyze the initial image, determine at least one image area contained in the initial image, and determine the focus information corresponding to each image area;

[0046] The focus determination module is used to determine a target image area from at least one image area and obtain target focus information corresponding to the target image area;

[0047] The shooting module is used to complete shooting based on the target focus information.

[0048] According to another aspect of the present invention, there is provided a computing device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;

[0049] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the above-mentioned video recording head snapshot and analysis method based on the AI ​​algorithm.

[0050] According to another aspect of the present invention, a computer storage medium is provided, wherein the storage medium stores at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the above-mentioned video camera snapshot and analysis method based on an AI algorithm.

[0051] According to the technical solution provided by the present invention, external image information is obtained from the camera as the initial image; the initial image is analyzed to determine at least one image area contained in the initial image, and the focus information corresponding to each image area is determined; the target image area is determined from at least one image area, and the target focus information corresponding to the target image area is obtained; and shooting is completed based on the target focus information. After the image is acquired through the camera, it is determined based on the preset judgment conditions whether the current image needs to be stored, and the current image is used as the initial image for further analysis, thereby making a judgment at the beginning of image acquisition, that is, whether the user has a shooting intention for this image, whether it needs to be set as the initial image and stored for subsequent analysis and shooting, intelligently analyzing the user's shooting intention, and reasonably screening the image information acquired by the image sensor, which greatly reduces the possibility of analyzing invalid images, improves the analysis and shooting efficiency, and saves shooting time; by constructing a neural network model, the image is divided after training, and the focus information is determined for the divided areas respectively, so that the subsequent process can quickly obtain the focus information for the image. The focus information is obtained by analyzing the initial image, thereby improving the shooting efficiency; the image features of the initial image are obtained, and the model is used to compare the image features in the historical image to determine a similar comparison image, and the focus method for the initial image is intelligently determined according to the focus method in the comparison image. Therefore, based on the historical data, the usual focus area and the focus method for similar images are scientifically determined, and the corresponding focus information obtained previously is further intelligently selected, thereby effectively improving the scientific nature of the focus selection for various shooting targets, making it more in line with the intentions of the usual shooting users, and greatly improving the shooting efficiency and user experience; based on the selected focus information, the focus and focal length are adjusted to complete the shooting, thereby speeding up the shooting process. In addition, by adjusting the attention weights in the image segmentation model, it is possible to more accurately screen out images with similar features from historical images for comparison, thereby improving the efficiency and accuracy of selection. The similarity between the selected images and the initial images is calculated to further complete the screening and improve the accuracy of image selection. Finally, the priority is determined by counting the number of various focus methods, and the selected focus method is determined accordingly, thereby making the selection of the focus method more intelligent and further improving the user experience.

[0052] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0053] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0055] Figure 1 A schematic diagram of a process of a video recording and analyzing method based on an AI algorithm according to an embodiment of the present invention is shown;

[0056] Figure 2 A schematic flow chart of a method for determining an initial image according to an embodiment of the present invention is shown;

[0057] Figure 3 A schematic flow chart of a method for determining focus information during a shooting process according to an embodiment of the present invention is shown;

[0058] Figure 4 A schematic flow chart of a method for determining a target image area in an image according to an embodiment of the present invention is shown;

[0059] Figure 5 The structure block diagram of a video recording and analysis system based on an AI algorithm according to an embodiment of the present invention is shown;

[0060] Figure 6 A schematic structural diagram of a computing device according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0061] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0062] Figure 1 The figure shows a flow chart of a method for video recording and analyzing based on an AI algorithm according to an embodiment of the present invention. The method comprises the following steps:

[0063] Step S101, obtaining external image information as an initial image through a camera.

[0064] Step S102 , analyzing the initial image, determining at least one image region included in the initial image, and determining focus information corresponding to each image region.

[0065] Specifically, an image segmentation model is constructed based on a neural network;

[0066] Using past historical images to train the image segmentation model, and obtaining a trained image segmentation model;

[0067] Analyzing the initial image based on the trained image segmentation model, and dividing the initial image into at least one image region;

[0068] For each image area, the corresponding focus information is determined.

[0069] Preferably, the image segmentation model can be based on an R-CNN (Region proposals-CNN, candidate region convolutional neural network) model, an LSTM (Long Short-Term Memory) model, a DINO (DETR with Improved deNoising AnchOr Boxes, improved end-to-end target detector for denoising anchor boxes) model, a U-Net (U-shaped Network) model or a SAM (Segment Anything Model) model, and the specific selection is not limited here.

[0070] Step S103: determine a target image area from at least one image area, and obtain target focus information corresponding to the target image area.

[0071] Step S104, completing the shooting based on the target focus information.

[0072] Specifically, obtaining target focus information;

[0073] Adjusting the focus and focal length of the initial image according to the target focus information to generate a focused image;

[0074] In response to the user's focus determination operation completing the photographing, the focused image is determined as a final photographed image, and is output and stored.

[0075] According to the video camera quick shooting and analysis method based on the AI ​​algorithm provided in this embodiment, the external image information is obtained from the camera as the initial image; the initial image is analyzed to determine at least one image area contained in the initial image, and the focus information corresponding to each image area is determined; the target image area is determined from at least one image area, and the target focus information corresponding to the target image area is obtained; and the shooting is completed based on the target focus information. Through the video camera quick shooting and analysis method based on the AI ​​algorithm provided in this embodiment, after the image is obtained by the camera, a neural network model is constructed, and the image is divided after the training is completed, and the focus information is determined for the divided areas respectively, so that the subsequent process can quickly obtain the focus information, thereby improving the shooting efficiency; by analyzing the initial image, the image features of the initial image are obtained, and the usual focus area and the focus method used for similar images are scientifically determined, and the target image area is further intelligently selected and the corresponding focus information is obtained, thereby effectively improving the scientific nature of the focus selection for various shooting targets, so that it can better meet the intention of the usual shooting user, and greatly improve the shooting efficiency and user experience; based on the selected focus information, the focus and focal length are adjusted to complete the shooting, which speeds up the shooting process.

[0076] Figure 2 A schematic flow chart of a method for determining an initial image according to an embodiment of the present invention is shown;

[0077] like Figure 2 As shown, the method comprises the following steps:

[0078] Step S201: receiving external image information based on a camera.

[0079] Step S202: store the received external image information and determine whether it meets a preset judgment condition.

[0080] Preferably, the preset judgment condition may be whether the retention time of the external image information in the viewfinder exceeds a preset time.

[0081] Among them, the preset time length can be set to 0.1s.

[0082] Specifically, if the preset judgment condition is met, step S203 is executed; if the preset judgment condition is not met, step S204 is executed.

[0083] Step S203: the external image information is set as the initial image.

[0084] Step S204: reusing the camera to obtain new external image information.

[0085] According to the above method, it is possible to determine whether the current image needs to be stored based on preset judgment conditions such as preset time length, and use the current image as the initial image for further analysis. Therefore, a judgment is made at the initial stage of acquiring the image, that is, whether the user has the intention to shoot this image, whether it needs to be set as the initial image and stored for subsequent analysis and shooting. The user's shooting intention is intelligently analyzed, and the image information obtained by the image sensor is reasonably screened, which greatly reduces the possibility of analyzing invalid images, improves analysis and shooting efficiency, and saves shooting time.

[0086] Figure 3 A schematic flow chart of a method for determining focus information during a shooting process according to an embodiment of the present invention is shown;

[0087] like Figure 3 As shown, the method comprises the following steps:

[0088] Step S301, analyzing and acquiring image features of an initial image.

[0089] Step S302: based on the image features of the initial image, a comparison image similar to the initial image is selected from the historical images.

[0090] Preferably, the image segmentation model is used to determine the image features of the initial image and the image features corresponding to each historical image, and the comparison image is obtained by screening based on the features.

[0091] Step S303: determining, according to the contrast image, an image region corresponding to the focus in the contrast image.

[0092] Step S304, determining the target image area in the initial image by analyzing the selection of the image area corresponding to the focus in the comparison image, and obtaining the target focus information corresponding to the target image area.

[0093] According to the above method, the image features of the initial image can be obtained by analyzing the initial image, and the image features in the historical image can be compared with the model to determine a similar comparison image. According to the focus method in the comparison image, the focus method for the initial image can be intelligently determined, thereby scientifically determining the usual focus area and focus method for similar images, thereby improving shooting efficiency and user experience.

[0094] Furthermore, Figure 4 A schematic flow chart of a method for determining a target image area in an image according to an embodiment of the present invention is shown;

[0095] like Figure 4 As shown, the method comprises the following steps:

[0096] Step S401 : determining a plurality of first images having the same image features from historical images based on an image segmentation model according to image features of an initial image.

[0097] Preferably, the image partition model is based on the SAM model, and the weight matrix in the image partition model is adjusted based on the low-rank matrix, thereby adjusting the Query parameter Q′ and the Value parameter V′.

[0098] Q′=X in (M Q +ΔM Q )=X in (M Q +A Q B Q )

[0099] V′=X in (M V +ΔM V )=X in (M V +A V B V )

[0100] Among them, X in is the input matrix; M Q is the query weight matrix; ΔM Q A is the query weight adjustment matrix; Q , B Q is the query low-rank matrix; M V is the numerical weight matrix; ΔM V A is the numerical weight adjustment matrix; V , B V is a numerical low-rank matrix;

[0101] An adjusted attention weight is derived based on the adjusted weight matrix, and then an adjusted image segmentation model is obtained. Based on the image features of the initial image, a plurality of first images having the same image features are determined from the historical images.

[0102] Step S402: Calculate the similarity between each first image and the initial image using the image segmentation model.

[0103] Preferably, for the plurality of first images, based on the image features of the initial image and the image features of the first image, the similarity between the two is calculated.

[0104]

[0105] Among them, S(t begin ,t sample ) is the similarity value; t begin is the image feature of the initial image; tsample is the image feature of the first image; d represents the dimension of the image feature; c represents the sequence number of the current dimension; is the component of the image feature of the initial image in the current dimension; is the component of the first image feature in the current dimension.

[0106] Step S403: determine the first image exceeding a preset similarity threshold as a comparison image.

[0107] Specifically, the similarity threshold can be set according to the actual situation of the user, and is not limited here. Generally speaking, if there are many historical images, or there are many shooting tasks of the same type and style, the similarity threshold can be increased accordingly.

[0108] Step S404: performing statistics according to the image areas corresponding to the focal points in each of the compared images to determine the number of images in which the focal points in the compared images are located in different image areas.

[0109] Step S405 : determining the priority of the image area corresponding to the focus based on the number of images, and determining the target image area in the initial image based on the image area with the highest priority.

[0110] Specifically, the priority value is the number of images whose focus corresponds to a certain type of specific image area. In the comparison image, the more images whose focus corresponds to a certain type of specific image area, the higher the priority.

[0111] According to the above method, by adjusting the attention weight in the image segmentation model, it is possible to more accurately screen out images with similar features from historical images for comparison, thereby improving the efficiency and accuracy of selection. In addition, for the selected image, the similarity between it and the initial image is calculated to further complete the screening and improve the accuracy of image selection. Finally, the priority is determined by counting the number of various focusing methods, and the selected focusing method is determined accordingly, thereby making the selection of the focusing method more intelligent and further improving the user experience.

[0112] Figure 5 The structure block diagram of a video recording and analysis system based on an AI algorithm according to an embodiment of the present invention is shown;

[0113] like Figure 5 As shown, the system includes: an image acquisition module 501, a feature analysis module 502, a focus determination module 503 and a shooting module 504; wherein,

[0114] The image acquisition module 501 is used to acquire external image information as an initial image through a camera.

[0115] Specifically, the image acquisition module 501 is further used to:

[0116] Receiving external image information based on a camera;

[0117] The received external image information is stored and whether it meets the preset judgment conditions is determined;

[0118] If yes, the external image information is set as the initial image; if no, the camera is reused to obtain new external image information.

[0119] The feature analysis module 502 is used to analyze the initial image, determine at least one image region included in the initial image, and determine focus information corresponding to each image region.

[0120] Specifically, the feature analysis module 502 is further used to:

[0121] Construct image segmentation model based on neural network;

[0122] Using past historical images to train the image segmentation model, and obtaining a trained image segmentation model;

[0123] Analyzing the initial image based on the trained image segmentation model, and dividing the initial image into at least one image region;

[0124] For each image area, the corresponding focus information is determined.

[0125] Preferably, the feature analysis module 502 is further used to:

[0126] According to the image features of the initial image, a plurality of first images having the same image features are determined from the historical images based on the image segmentation model;

[0127] Calculating the similarity between each first image and the initial image using the image segmentation model;

[0128] Determine the first image exceeding a preset similarity threshold as a comparison image;

[0129] According to the image areas corresponding to the focal points in each comparison image, statistics are performed to determine the number of images in which the focal points in the comparison image are located in different image areas;

[0130] Based on the number of images, the priority of the image area corresponding to the focus is determined, and the target image area in the initial image is determined based on the image area with the highest priority.

[0131] Preferably, the feature analysis module 502 is further used to:

[0132] The image partition model is based on the SAM model, and the weight matrix in the image partition model is adjusted based on the low-rank matrix, thereby adjusting the Query parameter Q′ and the Value parameter V′.

[0133] Q′=X in (M Q +ΔM Q )=X in (M Q +A Q B Q )

[0134] V′=X in (M V +ΔM V )=X in (M V +A V B V )

[0135] Among them, X in is the input matrix; M Q is the query weight matrix; ΔM Q A is the query weight adjustment matrix; Q , B Q is the query low-rank matrix; M V is the numerical weight matrix; ΔM V A is the numerical weight adjustment matrix; V , B V is a numerical low-rank matrix;

[0136] Based on the adjusted weight matrix, an adjusted attention weight is obtained, and then an adjusted image segmentation model is obtained, and based on the image features of the initial image, a plurality of first images having the same image features are determined from the historical images;

[0137] For the plurality of first images, based on the image features of the initial image and the image features of the first image, the similarity between the two is calculated,

[0138]

[0139] Among them, S(t begin ,t sample ) is the similarity value; t begin is the image feature of the initial image; t sample is the image feature of the first image; d represents the dimension of the image feature; c represents the sequence number of the current dimension; is the component of the image feature of the initial image in the current dimension; is the component of the first image feature in the current dimension.

[0140] The focus determination module 503 is used to determine a target image area from at least one image area, and obtain target focus information corresponding to the target image area.

[0141] Specifically, the focus determination module 503 is further configured to:

[0142] Analyze and obtain image features of the initial image;

[0143] Based on the image features of the initial image, a comparison image similar to the initial image is selected from the historical images;

[0144] According to the contrast image, determining an image area corresponding to a focus in the contrast image;

[0145] By analyzing the selection of the image area corresponding to the focus in the comparison image, the target image area in the initial image is determined, and the target focus information corresponding to the target image area is obtained.

[0146] The shooting module 504 is used to complete shooting based on the target focus information.

[0147] Specifically, the shooting module 504 is further used to:

[0148] Get target focus information;

[0149] Adjusting the focus and focal length of the initial image according to the target focus information to generate a focused image;

[0150] In response to the user's focus determination operation completing the photographing, the focused image is determined as a final photographed image, and is output and stored.

[0151] The video camera snapshot and analysis system based on the AI ​​algorithm provided in this embodiment includes: an image acquisition module, a feature analysis module, a focus determination module and a shooting module; wherein the image acquisition module is used to acquire external image information as an initial image according to a camera; the feature analysis module is used to analyze the initial image, determine at least one image area contained in the initial image, and determine the focus information corresponding to each image area; the focus determination module is used to determine a target image area from at least one image area, and acquire target focus information corresponding to the target image area; the shooting module is used to complete shooting based on the target focus information. Through the video camera quick shooting and analysis system based on AI algorithm provided by this embodiment, after acquiring the image through the camera, it is determined whether the current image needs to be stored based on the preset judgment conditions, and the current image is used as the initial image for further analysis, so that a judgment is made at the initial stage of acquiring the image, that is, whether the user has the intention to shoot this image, whether it needs to be set as the initial image and stored for subsequent analysis and shooting, and the user's shooting intention is intelligently analyzed, and the image information obtained by the image sensor is reasonably screened, which greatly reduces the possibility of analyzing invalid images, improves the analysis and shooting efficiency, and saves shooting time; by constructing a neural network model, the image is divided after training, and the focus information is determined for the divided areas respectively. information so that subsequent processes can quickly obtain focus information, thereby improving shooting efficiency; by analyzing the initial image, the image features of the initial image are obtained, and the model is used to compare the image features in the historical image to determine a similar comparison image, and the focus method for the initial image is intelligently determined according to the focus method in the comparison image, thereby scientifically determining the usual focus area and focus method for similar images based on historical data, and further intelligently selecting the corresponding focus information previously obtained, thereby effectively improving the scientific nature of focus selection for various shooting targets, making it more in line with the intentions of normal shooting users, and greatly improving shooting efficiency and user experience; based on the selected focus information, the focus and focal length are adjusted to complete the shooting, thereby speeding up the shooting process. In addition, by adjusting the attention weights in the image segmentation model, it is possible to more accurately screen out images with similar features from historical images for comparison, thereby improving the efficiency and accuracy of selection. The similarity between the selected images and the initial images is calculated to further complete the screening and improve the accuracy of image selection. Finally, the priority is determined by counting the number of various focus methods, and the selected focus method is determined accordingly, thereby making the selection of the focus method more intelligent and further improving the user experience.

[0152] The present invention also provides a non-volatile computer storage medium, which stores at least one executable instruction. The executable instruction can execute a video camera snapshot and analysis method based on an AI algorithm in any of the above method embodiments.

[0153] Figure 6 A schematic diagram of the structure of a computing device according to an embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the computing device.

[0154] like Figure 6 As shown, the computing device may include: a processor (processor) 602 , a communication interface (Communications Interface) 604 , a memory (memory) 606 , and a communication bus 608 .

[0155] in:

[0156] The processor 602 , the communication interface 604 , and the memory 606 communicate with each other via a communication bus 608 .

[0157] The communication interface 604 is used to communicate with other devices such as clients or other servers.

[0158] The processor 602 is used to execute the program 610, and specifically can execute the relevant steps in the above-mentioned embodiment of the video head snapshot shooting and analysis method based on the AI ​​algorithm.

[0159] Specifically, the program 610 may include program codes, and the program codes include computer operation instructions.

[0160] The processor 602 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the computing device may be processors of the same type, such as one or more CPUs; or may be processors of different types, such as one or more CPUs and one or more ASICs.

[0161] The memory 606 is used to store the program 610. The memory 606 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0162] Program 610 can be specifically used to enable processor 602 to execute a video head snapshot and analysis method based on an AI algorithm in any of the above method embodiments. The specific implementation of each step in program 610 can refer to the corresponding descriptions in the corresponding steps and units in the above-mentioned video head snapshot and analysis method embodiment based on an AI algorithm, which will not be repeated here. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described devices and modules can refer to the corresponding process description in the aforementioned method embodiment, which will not be repeated here.

[0163] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the present invention is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the present invention described herein, and the description of the above specific languages ​​is for disclosing the best mode of the present invention.

[0164] In the description provided herein, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0165] Similarly, it should be understood that in order to streamline the present disclosure and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the intention that the claimed invention requires more features than those expressly recited in each claim. Rather, as reflected in the claims, inventive aspects lie in less than all of the features of the individual embodiments previously disclosed. Therefore, the claims that follow the detailed description are hereby expressly incorporated into the detailed description, with each claim itself serving as a separate embodiment of the present invention.

[0166] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0167] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present invention and form different embodiments. For example, in the claims, any one of the claimed embodiments may be used in any combination.

[0168] The various component embodiments of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components in the embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., computer program and computer program product) for executing part or all of the methods described herein. Such a program implementing the present invention can be stored on a computer-readable medium, or can have the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0169] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A video camera quick-shot and analysis method based on AI algorithm, comprising: Acquire external image information according to the camera as an initial image; Analyze the initial image, determine at least one image area included in the initial image, and determine focus information corresponding to each image area; Determine a target image area from at least one image area, and obtain target focus information corresponding to the target image area; Shooting is completed based on the target focus information.

2. The method according to claim 1, characterized in that: The step of obtaining external image information as an initial image according to a camera further includes: Receiving external image information based on a camera; The received external image information is stored and whether it meets the preset judgment conditions is determined; If yes, the external image information is set as the initial image; if no, the camera is reused to obtain new external image information.

3. The method according to claim 1, characterized in that The analyzing the initial image, determining at least one image area contained in the initial image, and determining focus information corresponding to each image area, further includes: Construct image segmentation model based on neural network; Using past historical images to train the image segmentation model, and obtaining a trained image segmentation model; Analyzing the initial image based on the trained image segmentation model, and dividing the initial image into at least one image region; For each image area, the corresponding focus information is determined.

4. The method according to claim 1, characterized in that The step of determining a target image area from at least one image area and acquiring target focus information corresponding to the target image area further includes: Analyze and obtain image features of the initial image; Based on the image features of the initial image, a comparison image similar to the initial image is selected from the historical images; According to the contrast image, determining an image area corresponding to a focus in the contrast image; By analyzing the selection of the image area corresponding to the focus in the comparison image, the target image area in the initial image is determined, and the target focus information corresponding to the target image area is obtained.

5. The method according to claim 1, characterized in that: The method of completing shooting based on the target focus information further includes: Get target focus information; Adjusting the focus and focal length of the initial image according to the target focus information to generate a focused image; In response to the user's focus determination operation completing the photographing, the focused image is determined as a final photographed image, and is output and stored.

6. The method according to claim 1, characterized in that The step of determining a target image area from at least one image area and acquiring target focus information corresponding to the target image area further includes: According to the image features of the initial image, a plurality of first images having the same image features are determined from the historical images based on the image segmentation model; Calculating the similarity between each first image and the initial image using the image segmentation model; Determine the first image exceeding a preset similarity threshold as a comparison image; According to the image areas corresponding to the focal points in each comparison image, statistics are performed to determine the number of images in which the focal points in the comparison image are located in different image areas; Based on the number of images, the priority of the image area corresponding to the focus is determined, and the target image area in the initial image is determined based on the image area with the highest priority.

7. The method according to claim 6, characterized in that The method further comprises: The image partitioning model is based on the SAM model, and the weight matrix in the image partitioning model is adjusted based on the low-rank matrix, thereby adjusting the Query parameter Q ′ and Value parameter V′, Q′=X in (M Q +ΔM Q )=X in (M Q +A Q B Q ) V′=X in (M V +ΔM V )=X in (M V +A V B V ) Among them, X in is the input matrix; M Q is the query weight matrix; ΔM Q A is the query weight adjustment matrix; Q , B Q is the query low-rank matrix; M V is the numerical weight matrix; ΔM V A is the numerical weight adjustment matrix; V , B V is a numerical low-rank matrix; Based on the adjusted weight matrix, an adjusted attention weight is obtained, and then an adjusted image segmentation model is obtained, and based on the image features of the initial image, a plurality of first images having the same image features are determined from the historical images; For the plurality of first images, based on the image features of the initial image and the image features of the first image, the similarity between the two is calculated, Among them, S(t begin ,t sample ) is the similarity value; t begin is the image feature of the initial image; t sample is the image feature of the first image; d represents the dimension of the image feature; c represents the sequence number of the current dimension; is the component of the image feature of the initial image in the current dimension; is the component of the first image feature in the current dimension.

8. A video camera quick-shot and analysis system based on AI algorithm, comprising: Image acquisition module, feature analysis module, focus determination module and shooting module; wherein, The image acquisition module is used to acquire external image information as an initial image through a camera; The feature analysis module is used to analyze the initial image, determine at least one image area contained in the initial image, and determine the focus information corresponding to each image area; The focus determination module is used to determine a target image area from at least one image area and obtain target focus information corresponding to the target image area; The shooting module is used to complete shooting based on the target focus information.

9. A computing device comprising: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the video capture and analysis method based on an AI algorithm as described in any one of claims 1-7.

10. A computer storage medium, wherein at least one executable instruction is stored in the storage medium, and the executable instruction enables a processor to execute operations corresponding to a video camera snapshot and analysis method based on an AI algorithm as described in any one of claims 1 to 7.

Images