Acquisition method and system based on camera motion data set

Abstract

The invention relates to the field of computer vision, and discloses an acquisition method and system based on a camera motion dataset, and the method comprises the steps: firstly obtaining initial parameters of a camera, determining a shooting range track, analyzing a motion mode and a track trend of the shooting range track, obtaining a track feature sequence through track analysis, and calculating a motion speed value; then analyzing a shooting scene, extracting key scene elements after evaluation, determining an acquisition path, detecting and optimizing acquisition control parameters based on path interference terms, calculating an acquisition optimization ratio, and finally generating a data acquisition seed, querying detailed parameters and an acquisition identifier according to the ratio so as to generate a data acquisition scheme. The acquisition efficiency of the camera motion data can be improved.

Description

Technical Field

[0001] The invention relates to a camera motion data set-based acquisition method and system, and belongs to the field of computer vision. Background Art

[0002] In today's era of rapid development of digital imaging and intelligent vision technology, cameras, as key devices for obtaining image information, are constantly expanding their application scenarios, covering many fields such as autonomous driving, virtual reality, robot navigation, film and television production, etc.

[0003] At present, the traditional camera motion data collection method is relatively simple, mostly relying on manual operation of the camera to perform simple translation, rotation and other actions to obtain data. This method is extremely inefficient, and the collected data is seriously insufficient in diversity and complexity, and it is difficult to meet the growing technical needs. In addition, some collection methods based on specific scenes or devices can obtain a certain amount of data under specific conditions, but due to scene limitations and equipment compatibility issues, the collected data lacks universality and extensiveness, and cannot effectively support algorithm research and development in a variety of complex application scenarios. Therefore, a collection method based on camera motion datasets is needed to improve the efficiency of camera motion data collection. Summary of the invention

[0004] The present invention provides a method and system for collecting camera motion data sets, the main purpose of which is to improve the collection efficiency of camera motion data.

[0005] To achieve the above object, the present invention provides a method for collecting camera motion data sets, comprising: Acquire initial parameters corresponding to the motion camera, determine a shooting range trajectory corresponding to the motion camera based on the initial parameters, analyze a motion mode corresponding to the shooting range trajectory, and query a motion trajectory trend corresponding to the motion mode; Performing trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identifying trajectory feature parameters in the trajectory feature sequence, and calculating motion speed values ​​corresponding to the trajectory feature parameters; Based on the motion speed value, analyzing the shooting scene corresponding to the motion camera, performing scene evaluation on the shooting scene to obtain a scene evaluation value, and extracting key scene elements in the shooting scene based on the scene evaluation value; Based on the key scene elements, determining an acquisition path corresponding to the motion camera, detecting a path interference item corresponding to the acquisition path, optimizing an acquisition control parameter of the motion camera during data acquisition based on the path interference item, and calculating an acquisition optimization ratio corresponding to the acquisition control parameter; Based on the acquisition optimization ratio, a data acquisition seed corresponding to the motion camera is generated, detailed acquisition parameters corresponding to the data acquisition seed are queried, and an acquisition identifier corresponding to the detailed acquisition parameters is identified. Based on the acquisition identifier, a data acquisition plan corresponding to the motion camera is generated.

[0006] Optionally, querying the movement trajectory trend corresponding to the movement mode includes: extracting a pattern feature vector from the motion pattern; Identify a vector impact factor corresponding to the pattern feature vector; Determining a motion constraint condition corresponding to the motion mode based on the vector influence factor; extracting the dominant motion variables in the motion constraint conditions; Based on the dominant motion variable, a motion trajectory trend corresponding to the motion pattern is queried.

[0007] Optionally, performing trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence includes: Discretizing the motion trajectory trend to obtain a discrete trajectory point set; Calculating the trajectory change between adjacent trajectory points in the discrete trajectory point set; Constructing a trajectory change matrix corresponding to the trajectory change amount; Extracting core eigenvalues ​​in the trajectory change matrix; Based on the core feature value, trajectory analysis is performed on the motion trajectory trend to obtain a trajectory feature sequence.

[0008] Optionally, calculating the trajectory change amount between adjacent trajectory points in the discrete trajectory point set includes: The trajectory change between adjacent trajectory points in the discrete trajectory point set is calculated using the following formula: in, represents the trajectory change between adjacent trajectory points in the discrete trajectory point set, Represents the time node corresponding to the set of discrete trajectory points, Represents the number index of the three dimensions of the x, y, and z axes of the spatial dimension. Indicates at time node When , the value of the trajectory point in the kth dimension is, Indicates at time node When , the value of the trajectory point in the kth dimension is, Represents the time interval between adjacent time nodes, Represents the proportionality factor.

[0009] Optionally, calculating the motion speed value corresponding to the trajectory characteristic parameter includes: The motion speed value corresponding to the trajectory characteristic parameter is calculated using the following formula: in, represents the motion speed value corresponding to the trajectory characteristic parameter, Indicates the total number of distance dimension parameters corresponding to the trajectory feature parameters, Indicates the quantity index corresponding to the distance dimension parameter, Indicates The trajectory length value corresponding to the distance dimension parameter, Indicates The weight value corresponding to the distance dimension parameter, Indicates the total number of resistance dimension parameters corresponding to the trajectory characteristic parameters, Indicates the quantity index corresponding to the resistance dimension parameter, Indicates The quantitative value corresponding to the resistance dimension parameter.

[0010] Optionally, analyzing the shooting scene corresponding to the motion camera based on the motion speed value includes: Dividing the motion speed value into intervals to obtain a speed interval set; Identifying typical scene features of each interval in the speed interval set; Analyze the key screen requirements corresponding to the typical scene features; Based on the image requirements, determining shooting parameter requirements corresponding to the action camera; Based on the shooting parameter requirements, a shooting scene corresponding to the motion camera is analyzed.

[0011] Optionally, extracting key scene elements in the shooting scene based on the scene evaluation value includes: Analyze the scenario evaluation index corresponding to the scenario evaluation value; Query the indicator distribution corresponding to the scenario evaluation indicator; Based on the indicator distribution, determine the scene threshold range corresponding to the scene evaluation indicator; Based on the scene threshold range, key scene elements in the shooting scene are extracted.

[0012] Optionally, determining a collection path corresponding to the motion camera based on the key scene element includes: Identifying spatial position information of the key scene elements in the shooting scene; Extracting a position coordinate set from the spatial position information; Performing regional analysis on the position coordinate set to obtain a key coordinate region; Analyze the shooting focus corresponding to the key coordinate area; Based on the shooting focus, a collection path corresponding to the motion camera is determined.

[0013] Optionally, the optimizing the acquisition control parameters of the motion camera during data acquisition based on the path interference term includes: Identify the project type corresponding to the path interference item; Based on the project type, collecting initial control parameters of the motion camera in the collection path; Based on the initial control parameters, determining a performance optimization dimension of the motion camera in a control scenario; Querying a set of optimization strategies corresponding to the data in the performance optimization dimension; Extracting key parameter features of each sub-strategy data in the optimization strategy set; Based on the key parameter characteristics, the acquisition control parameters of the motion camera during the data acquisition process are optimized.

[0014] In order to solve the above problems, the present invention also provides a collection system based on a camera motion data set, the system comprising: A trend query module, used to obtain initial parameters corresponding to the motion camera, determine the shooting range trajectory corresponding to the motion camera based on the initial parameters, analyze the motion mode corresponding to the shooting range trajectory, and query the motion trajectory trend corresponding to the motion mode; A speed value calculation module is used to perform trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identify trajectory feature parameters in the trajectory feature sequence, and calculate the motion speed value corresponding to the trajectory feature parameter; an element extraction module, configured to analyze a shooting scene corresponding to the motion camera based on the motion speed value, perform scene evaluation on the shooting scene to obtain a scene evaluation value, and extract key scene elements in the shooting scene based on the scene evaluation value; an optimization ratio calculation module, used to determine the acquisition path corresponding to the motion camera based on the key scene elements, detect the path interference item corresponding to the acquisition path, optimize the acquisition control parameters of the motion camera during the data acquisition process based on the path interference item, and calculate the acquisition optimization ratio corresponding to the acquisition control parameters; A scheme generating module is used to generate a data acquisition seed corresponding to the motion camera based on the acquisition optimization ratio, query the detailed acquisition parameters corresponding to the data acquisition seed, identify the acquisition identifier corresponding to the detailed acquisition parameters, and generate a data acquisition scheme corresponding to the motion camera based on the acquisition identifier.

[0015] Compared with the problems described in the background technology, the present invention obtains the initial parameters corresponding to the motion camera, which can provide a basic basis for the subsequent determination of the shooting range trajectory, ensure that the trajectory planning is more accurate and reasonable, and help to accurately analyze the motion mode, so that the query of the trajectory trend is more targeted and effective. The present invention performs trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, which can extract key features from a complex motion trajectory, and can more clearly understand the motion details of the motion camera, such as the turning points of the motion, the speed change nodes, etc., to help identify abnormal situations in the motion mode, and provide strong support for optimizing the data collection plan and improving the shooting effect. Furthermore, the present invention analyzes the shooting scene corresponding to the motion camera based on the motion speed value, and can determine whether the camera movement speed is appropriate in a specific scene. It is suitable to ensure the stability and clarity of the picture, and it can also infer the scene characteristics according to the speed change, such as it is possible to capture dynamic scenes during fast movement, which helps to optimize the shooting plan and improve the shooting quality and effect. Furthermore, the present invention determines the acquisition path corresponding to the motion camera based on the key scene elements, which can greatly improve the pertinence and efficiency of shooting. It ensures that the camera accurately captures the elements that play a key role in the shooting effect and avoids invalid shooting. The reasonably planned acquisition path helps to present the best perspective of the scene and coherently display the key elements. Finally, the present invention generates the data acquisition seeds corresponding to the motion camera based on the acquisition optimization ratio, which can improve the efficiency of data acquisition. By reasonably planning the acquisition seeds, unnecessary repeated acquisition or missing important areas can be reduced, making the acquisition process more efficient and orderly. Therefore, the acquisition method and system based on the camera motion data set provided by the embodiment of the present invention can improve the acquisition efficiency of camera motion data. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a flow chart of a method for collecting a camera motion data set provided by an embodiment of the present invention; Figure 2 A schematic diagram of modules of a camera motion data set acquisition system provided by an embodiment of the present invention.

[0017] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0019] The embodiment of the present application provides a method for collecting camera motion data sets. The execution subject of the method for collecting camera motion data sets includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for collecting camera motion data sets based on the camera motion data set can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0020] Embodiment 1: Reference Figure 1 FIG. 1 is a flow chart of a method for collecting a camera motion data set provided by an embodiment of the present invention. In this embodiment, the method for collecting a camera motion data set includes: S1. Acquire initial parameters corresponding to a motion camera, determine a shooting range trajectory corresponding to the motion camera based on the initial parameters, analyze a motion mode corresponding to the shooting range trajectory, and query a motion trajectory trend corresponding to the motion mode.

[0021] The present invention obtains the initial parameters corresponding to the motion camera, which can provide a basic basis for the subsequent determination of the shooting range trajectory, ensure that the trajectory planning is more accurate and reasonable, help to accurately analyze the motion mode, and make the query of the trajectory trend more targeted and effective.

[0022] Among them, the motion camera refers to a movable camera device used to collect image or video data, which has the ability to work in a dynamic environment, such as being used to monitor the surrounding road conditions on an autonomous driving vehicle, assist navigation on a robot, or capture dynamic images in film and television shooting. The motion camera is usually equipped with a variety of sensors, such as accelerometers, gyroscopes, etc.; the initial parameters refer to a series of settings or measured values ​​that need to be obtained before starting to collect motion camera data. These parameters include but are not limited to the camera's installation position coordinates, initial orientation angle, lens focal length, frame rate, resolution, etc. The installation position coordinates determine the camera's starting point in space, the initial orientation angle clarifies the camera's initial shooting direction, the lens focal length affects the size and range of the image, the frame rate determines the number of images collected per unit time, and the resolution is related to the clarity of the collected image or video. Optionally, the acquisition of the initial parameters corresponding to the motion camera can be achieved through the camera's built-in sensors, such as: accelerometers, gyroscopes and other tools.

[0023] Furthermore, the present invention determines the shooting range trajectory corresponding to the motion camera based on the initial parameters, and analyzes the motion pattern corresponding to the shooting range trajectory, so as to accurately plan the shooting content, ensure that the collected images or videos meet the expected requirements, and provide effective materials for subsequent data processing.

[0024] The shooting range trajectory refers to the spatial range covered by the lens of the motion camera during the shooting process and the moving path of the coverage range over time, which comprehensively considers the initial parameters of the camera (such as installation position, orientation angle, etc.), motion state (translation, rotation, zoom, etc.) and shooting duration. For example, in a monitoring scene of an autonomous driving vehicle, the motion camera is installed at the front of the vehicle. Its shooting range trajectory can be that as the vehicle travels, it starts from an initial rectangular area in front of the front of the vehicle, gradually extends forward and swings left and right at a certain angle, forming a dynamic three-dimensional space trajectory; the motion mode It refers to the movement law and method followed by the motion camera during the shooting process, which reflects the characteristics and features of the camera movement, including translational movement (such as uniform linear translation, variable speed curve translation, etc.), rotational movement (such as uniform rotation around a fixed axis, non-uniform rotation, etc.), compound movement (complex movement in which translation and rotation are performed simultaneously), etc. Optionally, the determination of the shooting range trajectory corresponding to the motion camera can be achieved through a feature detection algorithm, such as SIFT, SURF and other algorithms; the analysis of the motion pattern corresponding to the shooting range trajectory can be achieved through a machine learning algorithm, such as a support vector machine, a decision tree and other algorithms.

[0025] Furthermore, the present invention helps to predict the movement direction of the camera in advance by querying the movement trajectory trend corresponding to the movement mode, providing a forward-looking decision-making basis for related applications (such as path planning in autonomous driving and lens scheduling in film and television shooting), making the operation more accurate and efficient.

[0026] The motion trajectory trend describes the possible development direction and changes of the future motion trajectory of the camera in the current motion mode, through the analysis of information such as the dominant motion variables, such as whether it is continuously accelerating linear motion, decelerating turning motion, or maintaining uniform circular motion.

[0027] As an embodiment of the present invention, the query of the motion trajectory trend corresponding to the motion pattern includes: extracting the pattern feature vector in the motion pattern; identifying the vector influence factor corresponding to the pattern feature vector; determining the motion constraint condition corresponding to the motion pattern based on the vector influence factor; extracting the dominant motion variable in the motion constraint condition; and querying the motion trajectory trend corresponding to the motion pattern based on the dominant motion variable.

[0028] Among them, the pattern feature vector refers to a mathematical vector that can quantitatively describe the characteristics of the motion pattern, which contains a variety of feature information related to the motion pattern, such as the speed, acceleration, direction change frequency, motion amplitude, etc. of the motion; the vector influencing factor refers to the factor in the pattern feature vector that plays a key role in determining the motion pattern and subsequent analysis. For example, in the feature vector describing the camera rotation motion pattern, the value of the rotation speed dimension and its rate of change over time are important vector influencing factors; the motion constraint condition refers to a series of conditions that limit the camera motion based on the vector influencing factor. These conditions are derived from the information contained in the vector influencing factor, including but not limited to the spatial range restriction of the motion, such as the camera can only rotate within a specific angle range; speed restriction, that is, the camera motion speed cannot exceed a certain threshold; and time restriction, such as completing a specific motion action within a specific time period; the dominant motion variable refers to the variable extracted from the motion constraint condition that plays a major role in determining the motion trajectory trend. It is the core element among many motion constraint conditions. For example, in the linear translation motion mode of the camera, the direction and speed of the translation are the dominant motion variables.

[0029] Further, the extraction of the pattern feature vector in the motion pattern can be achieved through a principal component analysis algorithm, such as: the data can be subjected to dimensionality reduction processing to obtain the pattern feature vector while retaining the main information; the identification of the vector influencing factor corresponding to the pattern feature vector can be achieved through a neural network method, such as: using a multi-layer perceptron (MLP), taking the pattern feature vector and possible influencing factors as the input layer, and obtaining weights related to the influencing factors at the output layer after nonlinear transformation of the hidden layer, and these weights can be used as vector influencing factors; the determination of the motion constraint corresponding to the motion pattern can be achieved through a CSP algorithm, such as: constructing the various variables and constraints of the motion pattern into a CSP model, and finding a solution that satisfies all constraints through a search algorithm, thereby determining the motion constraint; the extraction of the dominant motion variable in the motion constraint can be achieved through a particle swarm optimization algorithm, such as: the motion constraint can be converted into a fitness function, and the PSO algorithm can be used to search for the variable combination that makes the fitness optimal (i.e., the most dominant) to extract the dominant motion variable; the query of the motion trajectory trend corresponding to the motion pattern can be achieved through a cluster analysis method, such as: clustering similar motion trajectories, and then analyzing the center trajectory direction of each cluster to determine the overall motion trajectory trend.

[0030] S2. Perform trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identify trajectory feature parameters in the trajectory feature sequence, and calculate motion speed values ​​corresponding to the trajectory feature parameters.

[0031] The present invention obtains a trajectory feature sequence by performing trajectory analysis on the motion trajectory trend, and can extract key features from a complex motion trajectory, so as to provide a clearer insight into the motion details of the motion camera, such as the turning points of the motion, speed change nodes, etc., and help identify abnormal situations in the motion mode, thus providing strong support for optimizing the data collection scheme and improving the shooting effect.

[0032] Among them, the trajectory feature sequence refers to a series of ordered feature value combinations obtained after analyzing the motion trajectory trend based on the core feature value. It is a feature description of the entire motion trajectory and contains key information of the motion trajectory, such as the mode, frequency, amplitude and other characteristics of the motion.

[0033] As an embodiment of the present invention, the trajectory analysis of the motion trajectory trend to obtain a trajectory feature sequence includes: discretizing the motion trajectory trend to obtain a discrete trajectory point set; calculating the trajectory change amount between adjacent trajectory points in the discrete trajectory point set; constructing a trajectory change matrix corresponding to the trajectory change amount; extracting core eigenvalues ​​in the trajectory change matrix; and based on the core eigenvalues, performing trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence.

[0034] Among them, the discrete trajectory point set refers to a set of a series of discrete points obtained after dividing the continuous motion trajectory trend according to a certain time interval or space interval. For example, the position of the motion camera is recorded every 0.1 second, and these position points constitute the discrete trajectory point set; the trajectory change refers to the change value of the position, speed, acceleration, etc. between adjacent trajectory points in the discrete trajectory point set. For example, the displacement difference between two adjacent points represents the change in position, and the speed difference represents the change in speed; the trajectory change matrix refers to a matrix constructed with trajectory change as elements. For example, the first row of the matrix can represent the displacement change between adjacent points, and the second row represents the speed change, etc.; the core eigenvalue refers to a representative and important value extracted from the trajectory change matrix. These values ​​can highlight the main characteristics of the motion trajectory, such as the periodicity of the motion, the mutation point, etc.

[0035] Furthermore, the discretization of the motion trajectory trend can be achieved through an equidistant division method, such as dividing the range of the entire motion trajectory into a number of small rectangular areas at fixed distance intervals, and the vertex or center point of each area can be used as a discrete trajectory point; the calculation of the trajectory change between adjacent trajectory points in the discrete trajectory point set can be achieved through a numerical differentiation method, such as using the central difference method to calculate the speed and acceleration of each point, and then calculating the speed change and acceleration change between adjacent points, which together with the position change constitute the trajectory change; the construction of the trajectory change matrix corresponding to the trajectory change can be achieved through a matrix filling method Implementation, such as: the i-th row and the first column are filled with the position change between the i-th and i+1-th trajectory points, the i-th row and the second column are filled with the speed change, the row and the column are filled with the acceleration change, and finally a trajectory change matrix is ​​obtained; the extraction of the core eigenvalues ​​in the trajectory change matrix can be implemented by the eigenvalue decomposition method, such as: using the linalg.eigh function of the SciPy library in Python to perform eigenvalue decomposition on the matrix, obtain the eigenvalues ​​and sort them, and select the first few larger eigenvalues ​​as the core eigenvalues; the trajectory analysis of the motion trajectory trend can be implemented by a machine learning classification method, such as: support vector machine, decision tree and other algorithms.

[0036] As an embodiment of the present invention, the calculating of the trajectory change amount between adjacent trajectory points in the discrete trajectory point set includes: The trajectory change between adjacent trajectory points in the discrete trajectory point set is calculated using the following formula: in, represents the trajectory change between adjacent trajectory points in the discrete trajectory point set, Represents the time node corresponding to the set of discrete trajectory points, Represents the number index of the three dimensions of the x, y, and z axes of the spatial dimension. Indicates at time node When , the value of the trajectory point in the kth dimension is, Indicates at time node When , the value of the trajectory point in the kth dimension is, Represents the time interval between adjacent time nodes, Represents the proportionality factor.

[0037] Furthermore, the present invention can help accurately grasp the key information of the motion trajectory of the motion camera, such as special nodes of the motion, change rules, etc., by identifying the trajectory feature parameters in the trajectory feature sequence. These feature parameters also provide core data support for subsequent calculation of motion speed values, analysis of shooting scenes, etc., which helps to have a deeper understanding of the motion situation and lay a solid foundation for optimizing the data collection plan.

[0038] Among them, the trajectory characteristic parameters refer to the key parameters identified from the trajectory characteristic sequence that can describe the characteristics of the motion trajectory. These parameters contain characteristic information of the motion trajectory in terms of position, speed, acceleration, direction, etc. For example, the position coordinates of a specific point in the trajectory, the peak velocity, the rate of change of acceleration, the angle of change of the motion direction, etc., all belong to trajectory characteristic parameters, which can reflect the shape of the motion trajectory, the speed change of the motion, and the dynamic change trend of the trajectory, etc. Optionally, the identification of the trajectory characteristic parameters in the trajectory characteristic sequence can be achieved through parameter extraction tools, such as: MATLAB, Python and other tools.

[0039] Furthermore, the present invention can intuitively reflect the movement speed of the motion camera in a specific trajectory stage by calculating the motion speed value corresponding to the trajectory characteristic parameter, and can clearly know its dynamic changes, providing a key basis for analyzing the shooting scene, and helping to judge the rationality of the camera's movement in different scenes.

[0040] The motion speed value refers to a numerical value reflecting the speed of the motion camera under a specific motion trajectory, which comprehensively considers multiple factors such as distance dimension parameters and resistance dimension parameters. It is a quantitative motion speed representation used to describe the speed of the camera during movement.

[0041] As an embodiment of the present invention, the calculating the motion speed value corresponding to the trajectory characteristic parameter includes: The motion speed value corresponding to the trajectory characteristic parameter is calculated using the following formula: in, represents the motion speed value corresponding to the trajectory characteristic parameter, Indicates the total number of distance dimension parameters corresponding to the trajectory feature parameters, Indicates the quantity index corresponding to the distance dimension parameter, Indicates The trajectory length value corresponding to the distance dimension parameter, Indicates The weight value corresponding to the distance dimension parameter, Indicates the total number of resistance dimension parameters corresponding to the trajectory characteristic parameters, Indicates the quantity index corresponding to the resistance dimension parameter, Indicates The quantitative value corresponding to the resistance dimension parameter.

[0042] In detail, the distance dimension parameter refers to a type of parameter related to distance in the trajectory feature parameters, for example, in three-dimensional space, it corresponds to the distance-related quantities in the x, y, and z directions respectively, and may also include distance measurement parameters such as the arc length of the curve trajectory; the trajectory length value refers to the i-th distance dimension parameter, which is the length measurement value of the motion trajectory in this dimension. For example, in the x-axis direction, the straight-line distance or curve length that the camera travels from one point to another is the trajectory length value corresponding to the x-axis distance dimension parameter; the weight value refers to a coefficient corresponding to the i-th distance dimension parameter, which is used to measure the relative importance of the distance dimension parameter in calculating the motion speed value. The value range is generally between 0 and 1. If a distance dimension parameter has a greater impact on the motion speed, it can be given a higher weight. value; conversely, if the influence is small, the weight value is low; the resistance dimension parameter refers to a type of parameter in the trajectory characteristic parameters that is related to the resistance encountered during the movement. These parameters describe various resistance factors that may affect the movement speed, such as air resistance, ground friction, and the resistance caused by obstacles encountered during the movement; the quantized value refers to the jth resistance dimension parameter, which is the quantified value of the resistance factor. For example, for air resistance, a specific value may be calculated based on factors such as air density, the windward area of ​​the moving object, and speed to represent the size of the air resistance. This value is the quantized value corresponding to the resistance dimension parameter of air resistance.

[0043] S3. Analyze the shooting scene corresponding to the motion camera based on the motion speed value, perform scene evaluation on the shooting scene to obtain a scene evaluation value, and extract key scene elements in the shooting scene based on the scene evaluation value.

[0044] The present invention analyzes the shooting scene corresponding to the motion camera based on the motion speed value, and can determine whether the camera movement speed is appropriate in a specific scene to ensure a stable and clear picture. It can also infer scene features based on speed changes, such as whether a dynamic scene may be captured during rapid movement, thereby helping to optimize the shooting plan and improve shooting quality and effect.

[0045] The shooting scene refers to a specific environment and conditions suitable for sports camera shooting based on shooting parameter requirements, including the actual physical environment (such as indoors, outdoors, different terrains, etc.) and the camera settings, etc. For example, arranging the lights indoors and adjusting the camera position and angle according to the shooting parameter requirements constitutes a specific shooting scene.

[0046] As an embodiment of the present invention, analyzing the shooting scene corresponding to the motion camera based on the motion speed value includes: dividing the motion speed value into intervals to obtain a speed interval set; identifying typical scene features of each interval in the speed interval set; analyzing the picture requirements corresponding to the typical scene features; determining the shooting parameter requirements corresponding to the motion camera based on the picture requirements; and analyzing the shooting scene corresponding to the motion camera based on the shooting parameter requirements.

[0047] Among them, the speed interval set refers to a set formed by dividing the motion speed value into different intervals according to certain rules. For example, the speed value can be divided into a low-speed interval (such as 0-5 meters / second), a medium-speed interval (5-15 meters / second), a high-speed interval (15-30 meters / second), etc. These different speed intervals together constitute the speed interval set; the typical scene feature refers to the common scene characteristics corresponding to each interval in the speed interval set. For example, in the low-speed interval, the corresponding typical scene feature is shooting static objects or slowly moving objects, such as indoor still life shooting, slow movement of people, etc.; the medium-speed interval can correspond to shooting general dynamic scenes, such as normal vehicles on city streets, etc.; the high-speed interval can correspond to shooting fast-moving scenes, such as fast running in sports events, etc.; the picture demand focus refers to the aspects that need to be focused on when shooting the picture based on the typical scene features. For example, for static scenes corresponding to low-speed intervals, the focus of picture requirements may be high resolution to clearly present object details; for fast-moving scenes corresponding to high-speed intervals, the focus of picture requirements may be high frame rate to avoid image blur and ensure the continuity of moving objects; the shooting parameter requirements refer to the specific parameter setting requirements of the sports camera determined according to the focus of picture requirements, covering parameters such as aperture size, shutter speed, sensitivity (ISO), focal length, etc. For example, when the focus of picture requirements is high frame rate, the shooting parameter requirements may be to set a higher shutter speed and appropriate sensitivity to ensure the quality and stability of the picture during fast shooting.

[0048] Furthermore, the interval division of the motion speed value can be achieved by an equidistant division method, such as: if the motion speed value range is 0-50 meters / second, and the interval is set to 10 meters / second, then the speed intervals of 0-10, 10-20, 20-30, 30-40, and 40-50 can be divided to form a speed interval set; the identification of typical scene features of each interval in the speed interval set can be achieved by a machine learning classification method, such as: support vector machine, random forest and other classification algorithms; the analysis of the picture requirements corresponding to the typical scene features can be achieved by a decision tree model, such as: through training The decision tree model is trained. When a new typical scene feature is input, the corresponding picture requirement focus can be quickly determined. The determination of the shooting parameter requirements corresponding to the sports camera can be achieved through a mapping relationship establishment method, such as: by establishing a mapping table between the picture requirement focus and the shooting parameter, the corresponding shooting parameter requirements are determined according to the picture requirement focus obtained previously. The analysis of the shooting scene corresponding to the sports camera can be achieved through project management tools, such as Jira and other tools, and the results of each link are used as task input. By setting task dependencies and workflows, the shooting scene corresponding to the sports camera is finally summarized.

[0049] The present invention performs scene evaluation on the shooting scene to obtain a scene evaluation value, which can systematically measure the pros and cons of the shooting scene and clarify whether the current scene meets the shooting target. Through the evaluation value, problems in the scene, such as insufficient light, cluttered background, etc., can be quickly discovered, and targeted adjustments and optimizations can be made.

[0050] Among them, the scene evaluation value refers to a quantitative value, which is used to comprehensively reflect the suitability of the shooting scene for achieving the expected shooting effect. It is obtained by scoring and calculating multiple key elements in the shooting scene, such as light conditions, spatial layout, background complexity, color matching, etc. according to specific evaluation standards and weights. A higher scene evaluation value means that the scene performs well in terms of sufficient light, space conducive to framing, simple background and coordinated colors, and is more conducive to obtaining high-quality shooting images; otherwise, it indicates that there are some areas in the scene that need to be improved or adjusted to meet shooting requirements. Optionally, the scene evaluation of the shooting scene can be achieved through scene evaluation methods, such as: hierarchical analysis method, fuzzy comprehensive evaluation method and other methods.

[0051] Furthermore, the present invention extracts key scene elements in the shooting scene based on the scene evaluation value, can focus on the core content, can be optimized and highlighted in a targeted manner, and is conducive to the effective organization and analysis of data, so that it can automatically adjust shooting parameters and strategies according to key elements to better adapt to scene requirements.

[0052] Among them, the key scene elements refer to specific objects or parts in the shooting scene that play an important role in the shooting effect. For example, in an area that meets the threshold requirements of light intensity, color contrast, etc., there may be some elements with unique visual effects or closely related to the shooting subject, such as prominent main objects, guiding lines, special light and shadow effects, etc.

[0053] As an embodiment of the present invention, extracting key scene elements in the shooting scene based on the scene evaluation value includes: analyzing a scene evaluation indicator corresponding to the scene evaluation value; querying an indicator distribution corresponding to the scene evaluation indicator; determining a scene threshold range corresponding to the scene evaluation indicator based on the indicator distribution; and extracting key scene elements in the shooting scene based on the scene threshold range.

[0054] Among them, the scene evaluation index refers to the specific parameters or factors used to measure the quality and suitability of the shooting scene. These indicators cover multiple aspects of the shooting scene, such as light intensity, color contrast, background complexity, rationality of spatial layout, etc. Each indicator reflects the characteristics of the shooting scene from a specific angle and is the basis for quantitative evaluation of the shooting scene; the indicator distribution refers to the numerical distribution state of each scene evaluation index in the entire shooting scene, which describes the changes of each indicator in different areas or conditions. For example, for the light intensity index, its distribution can show which areas in the shooting scene are brighter and which areas are darker; for the color contrast index, the distribution can reflect the high and low distribution of contrast between different color areas; the scene threshold range refers to a numerical interval determined according to the indicator distribution and shooting requirements. When the value of a certain indicator falls within this range, it indicates that the scene condition corresponding to the indicator meets certain shooting requirements or standards. For example, for a scene of shooting close-ups of people, the scene threshold range of light intensity may be set in a specific interval to ensure that the light on the face of the person is clear and without shadows.

[0055] Furthermore, the analysis of the scene evaluation index corresponding to the scene evaluation value can be achieved through a feature selection algorithm, such as: chi-square test, mutual information method and other algorithms; the query of the indicator distribution corresponding to the scene evaluation index can be achieved through a data analysis tool, such as: LabVIEW, MATLAB and other tools; the determination of the scene threshold range corresponding to the scene evaluation index can be achieved through a machine learning threshold determination method, such as: using the SVM algorithm in the scikit-learn library, after learning a large amount of shooting scene data, determining the appropriate threshold range of indicators such as light uniformity and background complexity; the extraction of key scene elements in the shooting scene can be achieved through a deep learning framework, such as: TensorFlow, PyTorch and other frameworks.

[0056] S4. Based on the key scene elements, determine the acquisition path corresponding to the motion camera, detect the path interference item corresponding to the acquisition path, optimize the acquisition control parameters of the motion camera during the data acquisition process based on the path interference item, and calculate the acquisition optimization ratio corresponding to the acquisition control parameters.

[0057] The present invention determines the acquisition path corresponding to the motion camera based on the key scene elements, which can greatly improve the pertinence and efficiency of shooting. It ensures that the camera accurately captures the elements that play a key role in the shooting effect and avoids invalid shooting. A reasonably planned acquisition path helps to present the best viewing angle of the scene and coherently display the key elements.

[0058] The acquisition path refers to the movement route planned by the motion camera in the shooting scene in order to comprehensively and effectively capture key scene elements. For example, when shooting a large building, the acquisition path may move in a circle around the building to capture key parts of the building from different angles, such as the unique architectural appearance, exquisite decorative details, etc.

[0059] As an embodiment of the present invention, determining the acquisition path corresponding to the motion camera based on the key scene element includes: identifying the spatial position information of the key scene element in the shooting scene; extracting a position coordinate set in the spatial position information; performing regional analysis on the position coordinate set to obtain a key coordinate area; analyzing the shooting focus corresponding to the key coordinate area; and determining the acquisition path corresponding to the motion camera based on the shooting focus.

[0060] The spatial position information refers to the specific position description of the key scene element in the three-dimensional space of the entire shooting scene. For example, in an indoor shooting scene, a lamp with a unique shape is used as a key scene element, and its spatial position information can be described as 2 meters from the ground, 3 meters from the left wall, 4 meters from the rear wall, etc.; the position coordinate set refers to the set formed by quantizing the spatial position information of the key scene element in the form of coordinates. For example, in a two-dimensional plane scene, the position coordinates of a key scene element may be expressed as (x, y), a plurality of such coordinate values ​​constitute a position coordinate set; the key coordinate area refers to a specific area divided after performing regional analysis on the position coordinate set. For example, when shooting a sports event, the area where athletes gather and move around will be divided into a key coordinate area after analyzing the position coordinate set of relevant key scene elements (athletes, balls, etc.); the shooting focus refers to the focus direction determined according to the characteristics of the key coordinate area and the expected target of the shooting. Different key coordinate areas may have different characteristics. For example, some areas are rich in elements and have large dynamic changes, while some areas are brightly colored and have high contrast. The shooting focus may include highlighting the main elements, capturing dynamic moments, and showing color levels.

[0061] Furthermore, the identification of the spatial position information of the key scene elements in the shooting scene can be achieved through a convolutional neural network, such as YOLO, Faster R-CNN, etc.; the extraction of the position coordinate set in the spatial position information can be achieved through a coordinate conversion algorithm, such as two-dimensional coordinate conversion, three-dimensional coordinate conversion and other algorithms; the regional analysis of the position coordinate set can be achieved through a clustering algorithm, such as K-Means, DBSCAN and other tools; the analysis of the shooting focus corresponding to the key coordinate area can be achieved through a rule-based reasoning method, such as: these rules can be judged and reasoned by writing a Python program to obtain the shooting focus corresponding to each key coordinate area; the determination of the acquisition path corresponding to the motion camera can be achieved through a Dijkstra algorithm, such as: setting each key coordinate area as a node in the graph, and using the Dijkstra algorithm to start from the starting node and calculate the shortest path to all other nodes, thereby obtaining the acquisition path of the motion camera.

[0062] By detecting the path interference items corresponding to the acquisition path, the present invention can discover in advance the factors that may hinder the smooth movement of the motion camera, such as obstacles, signal interference sources, etc., so as to adjust the path planning in time to avoid shooting interruptions or problems such as shaking and blurring of the picture. By effectively avoiding interference items, the present invention can ensure the smooth progress of the acquisition work, improve the shooting efficiency, and provide a strong guarantee for obtaining high-quality image materials.

[0063] Among them, the path interference items refer to various factors that may have an adverse effect on the shooting process of the motion camera when it runs along the established acquisition path. These factors cover a wide range, including physical obstacles, such as pedestrians and branches that suddenly appear on the acquisition path, which will block the movement trajectory of the camera; there are also signal interference sources, such as strong electromagnetic fields, wireless signal transmitting equipment, etc., which can interfere with the camera's signal transmission and control, resulting in abnormal shooting parameters, data loss and other problems; in addition, unstable environmental factors, such as strong winds and uneven ground, also belong to path interference items. Optionally, the detection of the path interference items corresponding to the acquisition path can be achieved through a sensor-based detection method, such as: a laser radar constructs three-dimensional point cloud data of the surrounding environment by emitting a laser beam to the surroundings and measuring the time it takes for the laser to reflect back to the sensor, and can detect interference items such as obstacles on the acquisition path by analyzing the point cloud data.

[0064] Furthermore, the present invention optimizes the acquisition control parameters of the motion camera during data acquisition based on the path interference term, which can significantly improve the shooting effect. Through targeted adjustments, the influence of interference on image quality can be effectively avoided, such as adjusting the exposure parameters under light interference to ensure a clear picture, prevent data loss or abnormality due to interference, and allow the acquisition work to proceed smoothly to obtain high-quality image data.

[0065] Among them, the acquisition control parameters refer to various parameters that can be adjusted by the motion camera during the data acquisition process. These parameters directly affect the shooting effect and data quality, including the shutter speed, aperture size, ISO sensitivity, frame rate, etc. mentioned above, as well as parameters such as white balance and focus mode.

[0066] As an embodiment of the present invention, the optimization of the acquisition control parameters of the motion camera during the data acquisition process based on the path interference item includes: identifying the project type corresponding to the path interference item; based on the project type, acquiring the initial control parameters of the motion camera in the acquisition path; based on the initial control parameters, determining the performance optimization dimension of the motion camera in the control scenario; querying the optimization strategy set corresponding to the data in the performance optimization dimension; extracting the key parameter characteristics of each sub-strategy data in the optimization strategy set; and optimizing the acquisition control parameters of the motion camera during the data acquisition process based on the key parameter characteristics.

[0067] Among them, the project type refers to the specific category after classifying the path interference items, which is used to clarify the nature of different interference items, such as dividing the path interference items into categories such as physical obstacle interference, light interference, signal interference, etc.; the initial control parameters refer to the parameter setting values ​​of the motion camera before optimizing the current path interference items in the acquisition path. These parameters cover many aspects of camera shooting, such as shutter speed, aperture size, ISO sensitivity, frame rate, etc.; the performance optimization dimension refers to the direction in which the camera performance needs to be optimized based on the initial control parameters and path interference items. It is a classification and detailed description of the camera performance. For example, when facing light interference, the performance optimization dimension can be the image Brightness uniformity, color reproduction, etc.; when facing interference from physical obstacles, it can be dimensions such as the camera's focus speed and viewing angle range; the optimization strategy set refers to a set of feasible optimization methods developed for different performance optimization dimensions. For example, for the performance optimization dimension of image brightness uniformity, the optimization strategy set can include automatic exposure adjustment strategy, local brightness compensation strategy, etc.; the key parameter features refer to representative and decisive parameter characteristics extracted from the data of each sub-strategy in the optimization strategy set. These features can reflect the core points of each sub-strategy, such as parameter features such as the range of shutter speed adjustment and the amplitude of aperture value change in a certain optimization strategy.

[0068] Furthermore, the identification of the item type corresponding to the path interference item can be achieved through a rule-based classification method, such as: if the interference item is a physical entity and has a large volume, it can be classified as a "physical obstacle"; if the interference item causes a significant change in light intensity, it is classified as "light interference"; if the interference item affects signal transmission, it is classified as "signal interference"; the acquisition of the initial control parameters of the motion camera in the acquisition path can be achieved through a camera API call, such as: by sending a GET request to the API of the GoPro camera, obtaining the camera's status information, and extracting the camera's setting information from the returned JSON data as the initial control parameters; the determination of the performance optimization dimension of the motion camera in the control scenario can be achieved through a virtual scene simulation tool, such as: Unity, Unreal Engine and other tools; the query of the optimization strategy set corresponding to the data in the performance optimization dimension can be implemented through a database query system, such as: establishing a database containing various performance optimization dimensions and their corresponding optimization strategies, and querying the database to obtain the corresponding optimization strategy set by inputting the performance optimization dimension; the extraction of key parameter features of each sub-strategy data in the optimization strategy set can be implemented through a feature extraction algorithm, such as: for the automatic exposure adjustment algorithm, the key parameter features can include the exposure time adjustment step, the brightness threshold, etc.; for the manual exposure curve setting, the key parameter features can be the exposure value in different brightness ranges; the optimization of the acquisition control parameters of the motion camera during the data acquisition process can be implemented through a rule-based parameter adjustment algorithm, such as: if it is determined that the current ambient light is dim and the exposure time adjustment step of the automatic exposure adjustment algorithm is 0.1 second, the exposure time of the motion camera is adjusted from 1 / 60 second to 1 / 30 second, and the ISO value is adjusted to 400 according to the brightness threshold, thereby optimizing the acquisition control parameters.

[0069] S5. Based on the acquisition optimization ratio, generate a data acquisition seed corresponding to the motion camera, query detailed acquisition parameters corresponding to the data acquisition seed, identify an acquisition identifier corresponding to the detailed acquisition parameters, and generate a data acquisition plan corresponding to the motion camera based on the acquisition identifier.

[0070] The present invention generates data collection seeds corresponding to the motion camera based on the collection optimization ratio, which can improve the efficiency of data collection, reduce unnecessary repeated collection or omission of important areas through reasonable planning of collection seeds, and make the collection process more efficient and orderly.

[0071] Among them, the data collection seed refers to a starting element or data set that provides basic guidance and key information for motion camera data collection. It is generated based on factors such as the collection optimization ratio, and contains key information such as a specific collection location, time point, shooting parameter preference, etc. It is like a "guidance seed" for data collection work, which can guide the motion camera to collect data in a more optimized and targeted manner, laying the foundation for subsequent acquisition of high-quality data that meets specific needs. Optionally, the generation of the data collection seed corresponding to the motion camera can be achieved through a seed generation algorithm, such as: genetic algorithm, simulated annealing algorithm and other algorithms.

[0072] Furthermore, the present invention can significantly improve the standardization and efficiency of data collection by querying the detailed collection parameters corresponding to the data collection seed and identifying the collection identifier corresponding to the detailed collection parameters, allowing the motion camera to accurately perform collection tasks and avoid collection deviations caused by unknown parameters, thereby laying a solid foundation for the effective use of data.

[0073] The detailed acquisition parameters refer to a series of specific and detailed parameter sets set for the data acquisition work of the motion camera. These parameters cover all key aspects of camera shooting and are used to accurately guide the camera on how to collect data. For example, it includes shooting resolution settings, such as the common 1080p, 4K, etc. Different resolutions determine the clarity and detail richness of the image or video; frame rate parameters, such as 24 frames per second, 60 frames per second, etc., the frame rate affects the smoothness of the video; there are also parameters such as exposure time, sensitivity (ISO), white balance, etc.; the acquisition identifier refers to a specific symbol or code set to distinguish and identify different acquisition tasks or data sets, which can be By assigning a unique collection identifier to each collection task in the form of numbers, letters, strings or combinations thereof, data from different sources and for different purposes can be clearly distinguished. For example, in a complex environmental monitoring project, different collection identifiers are set for collection tasks in different areas and at different times. Optionally, the query of the detailed collection parameters corresponding to the data collection seed can be implemented by querying a relational database, such as MySQL, Oracle, etc.; the identification of the collection identifier corresponding to the detailed collection parameters can be implemented by a hash function method, such as using a hash function to process the detailed collection parameters to generate a unique hash value as a collection identifier.

[0074] The present invention generates a data collection scheme corresponding to the motion camera based on the collection identifier, which can greatly improve the pertinence and systematicness of the collection work. The collection identifier is used as a key index to accurately match relevant information, ensure that the collection scheme fully meets specific scenarios and needs, make the data collection process more standardized and orderly, and improve the quality and availability of the final collected data.

[0075] Among them, the data collection plan refers to a set of comprehensive plans carefully formulated for sports cameras, aiming to ensure efficient and accurate acquisition of required data, which specifies in detail the key elements of the sports camera in the data collection process, including but not limited to: clarifying the goal of the collection task, such as whether it is used for environmental monitoring, behavior analysis or other specific purposes; defining the scope of collection, such as a specific area, time period or a specific type of scene; accurately setting various parameters of the camera, such as resolution, frame rate, sensitivity, white balance, etc., to meet different shooting needs; planning the collection path to guide the movement trajectory of the camera in different spatial positions; setting the trigger conditions for data collection, such as the occurrence of a specific event, reaching a certain time point or meeting specific environmental parameters; it also covers the storage method, data format and preliminary data sorting and screening strategies after data collection. Optionally, the generation of the data collection plan corresponding to the sports camera can be achieved through a plan generation tool, such as Trello, Asana and other tools.

[0076] Compared with the problems described in the background technology, the present invention obtains the initial parameters corresponding to the motion camera, which can provide a basic basis for the subsequent determination of the shooting range trajectory, ensure that the trajectory planning is more accurate and reasonable, and help to accurately analyze the motion mode, so that the query of the trajectory trend is more targeted and effective. The present invention performs trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, which can extract key features from a complex motion trajectory, and can more clearly understand the motion details of the motion camera, such as the turning points of the motion, the speed change nodes, etc., to help identify abnormal situations in the motion mode, and provide strong support for optimizing the data collection plan and improving the shooting effect. Furthermore, the present invention analyzes the shooting scene corresponding to the motion camera based on the motion speed value, and can determine whether the camera movement speed is appropriate in a specific scene. It is suitable to ensure the stability and clarity of the picture, and it can also infer the scene characteristics according to the speed change, such as it is possible to capture dynamic scenes during fast movement, which helps to optimize the shooting plan and improve the shooting quality and effect. Furthermore, the present invention determines the acquisition path corresponding to the motion camera based on the key scene elements, which can greatly improve the pertinence and efficiency of shooting. It ensures that the camera accurately captures the elements that play a key role in the shooting effect and avoids invalid shooting. The reasonably planned acquisition path helps to present the best perspective of the scene and coherently display the key elements. Finally, the present invention generates the data acquisition seeds corresponding to the motion camera based on the acquisition optimization ratio, which can improve the efficiency of data acquisition. By reasonably planning the acquisition seeds, unnecessary repeated acquisition or missing important areas can be reduced, making the acquisition process more efficient and orderly. Therefore, the acquisition method and system based on the camera motion data set provided by the embodiment of the present invention can improve the acquisition efficiency of camera motion data.

[0077] Embodiment 2: like Figure 2 The figure shows a functional module diagram of a camera motion data set acquisition system according to the present invention.

[0078] The camera motion data set acquisition system 200 described in the present invention can be installed in an electronic device. According to the functions implemented, the camera motion data set acquisition system can include a trend query module 201, a speed value calculation module 202, an element extraction module 203, an optimization ratio calculation module 204 and a solution generation module 205. The module described in the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0079] In the embodiment of the present invention, the functions of each module / unit are as follows: The trend query module 201 is used to obtain initial parameters corresponding to the motion camera, determine the shooting range trajectory corresponding to the motion camera based on the initial parameters, analyze the motion mode corresponding to the shooting range trajectory, and query the motion trajectory trend corresponding to the motion mode; The speed value calculation module 202 is used to perform trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identify trajectory feature parameters in the trajectory feature sequence, and calculate the motion speed value corresponding to the trajectory feature parameter; The element extraction module 203 is used to analyze the shooting scene corresponding to the motion camera based on the motion speed value, perform scene evaluation on the shooting scene to obtain a scene evaluation value, and extract key scene elements in the shooting scene based on the scene evaluation value; The optimization ratio calculation module 204 is used to determine the acquisition path corresponding to the motion camera based on the key scene elements, detect the path interference item corresponding to the acquisition path, optimize the acquisition control parameters of the motion camera during the data acquisition process based on the path interference item, and calculate the acquisition optimization ratio corresponding to the acquisition control parameters; The solution generation module 205 is used to generate a data acquisition seed corresponding to the motion camera based on the acquisition optimization ratio, query the detailed acquisition parameters corresponding to the data acquisition seed, and identify the acquisition identifier corresponding to the detailed acquisition parameters, and generate a data acquisition solution corresponding to the motion camera based on the acquisition identifier.

[0080] In detail, the modules in the camera motion data set acquisition system 200 described in the embodiment of the present invention are used in the same manner as described above. Figure 1 The same technical means as the camera motion data set acquisition method described in , and can produce the same technical effects, will not be repeated here.

[0081] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for collecting camera motion data sets, characterized in that: The method comprises: Acquire initial parameters corresponding to the motion camera, determine a shooting range trajectory corresponding to the motion camera based on the initial parameters, analyze a motion mode corresponding to the shooting range trajectory, and query a motion trajectory trend corresponding to the motion mode; Performing trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identifying trajectory feature parameters in the trajectory feature sequence, and calculating motion speed values ​​corresponding to the trajectory feature parameters; Based on the motion speed value, analyzing the shooting scene corresponding to the motion camera, performing scene evaluation on the shooting scene to obtain a scene evaluation value, and extracting key scene elements in the shooting scene based on the scene evaluation value; Based on the key scene elements, determining an acquisition path corresponding to the motion camera, detecting a path interference item corresponding to the acquisition path, optimizing an acquisition control parameter of the motion camera during data acquisition based on the path interference item, and calculating an acquisition optimization ratio corresponding to the acquisition control parameter; Based on the acquisition optimization ratio, a data acquisition seed corresponding to the motion camera is generated, detailed acquisition parameters corresponding to the data acquisition seed are queried, and an acquisition identifier corresponding to the detailed acquisition parameters is identified. Based on the acquisition identifier, a data acquisition plan corresponding to the motion camera is generated.

2. The camera motion data set acquisition method according to claim 1, characterized in that: The querying of the movement trajectory trend corresponding to the movement mode includes: extracting a pattern feature vector from the motion pattern; Identify a vector impact factor corresponding to the pattern feature vector; Determining a motion constraint condition corresponding to the motion mode based on the vector influence factor; extracting the dominant motion variables in the motion constraint conditions; Based on the dominant motion variable, a motion trajectory trend corresponding to the motion pattern is queried.

3. The camera motion data set acquisition method according to claim 1, characterized in that: The performing trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence includes: Discretizing the motion trajectory trend to obtain a discrete trajectory point set; Calculating the trajectory change between adjacent trajectory points in the discrete trajectory point set; Constructing a trajectory change matrix corresponding to the trajectory change amount; Extracting core eigenvalues ​​in the trajectory change matrix; Based on the core feature value, trajectory analysis is performed on the motion trajectory trend to obtain a trajectory feature sequence.

4. The camera motion data set acquisition method according to claim 3, characterized in that: The calculating the trajectory change amount between adjacent trajectory points in the discrete trajectory point set includes: The trajectory change between adjacent trajectory points in the discrete trajectory point set is calculated using the following formula: in, represents the trajectory change between adjacent trajectory points in the discrete trajectory point set, Represents the time node corresponding to the set of discrete trajectory points, Represents the number index of the three dimensions of the x, y, and z axes of the spatial dimension. Indicates at time node When , the value of the trajectory point in the kth dimension is, Indicates at time node When , the value of the trajectory point in the kth dimension is, Represents the time interval between adjacent time nodes, Represents the proportionality factor.

5. The camera motion data set acquisition method according to claim 1, characterized in that: The calculating the motion speed value corresponding to the trajectory characteristic parameter includes: The motion speed value corresponding to the trajectory characteristic parameter is calculated using the following formula: in, represents the motion speed value corresponding to the trajectory characteristic parameter, Indicates the total number of distance dimension parameters corresponding to the trajectory feature parameters, Indicates the quantity index corresponding to the distance dimension parameter, Indicates The trajectory length value corresponding to the distance dimension parameter, Indicates The weight value corresponding to the distance dimension parameter, Indicates the total number of resistance dimension parameters corresponding to the trajectory characteristic parameters, Indicates the quantity index corresponding to the resistance dimension parameter, Indicates The quantitative value corresponding to the resistance dimension parameter.

6. The camera motion data set acquisition method according to claim 1, characterized in that: The analyzing the shooting scene corresponding to the motion camera based on the motion speed value includes: Dividing the motion speed value into intervals to obtain a speed interval set; Identifying typical scene features of each interval in the speed interval set; Analyze the key screen requirements corresponding to the typical scene features; Based on the image requirements, determining shooting parameter requirements corresponding to the action camera; Based on the shooting parameter requirements, a shooting scene corresponding to the motion camera is analyzed.

7. The camera motion data set acquisition method according to claim 1, characterized in that: The step of extracting key scene elements from the shooting scene based on the scene evaluation value includes: Analyze the scenario evaluation index corresponding to the scenario evaluation value; Query the indicator distribution corresponding to the scenario evaluation indicator; Based on the indicator distribution, determine the scene threshold range corresponding to the scene evaluation indicator; Based on the scene threshold range, key scene elements in the shooting scene are extracted.

8. The camera motion data set acquisition method according to claim 1, characterized in that: The determining, based on the key scene elements, a collection path corresponding to the motion camera includes: Identifying spatial position information of the key scene elements in the shooting scene; Extracting a position coordinate set from the spatial position information; Performing regional analysis on the position coordinate set to obtain a key coordinate region; Analyze the shooting focus corresponding to the key coordinate area; Based on the shooting focus, a collection path corresponding to the motion camera is determined.

9. The camera motion data set acquisition method according to claim 1, characterized in that: The optimizing the acquisition control parameters of the motion camera during the data acquisition process based on the path interference term includes: Identify the project type corresponding to the path interference item; Based on the project type, collecting initial control parameters of the motion camera in the collection path; Based on the initial control parameters, determining a performance optimization dimension of the motion camera in a control scenario; Querying a set of optimization strategies corresponding to the data in the performance optimization dimension; Extracting key parameter features of each sub-strategy data in the optimization strategy set; Based on the key parameter characteristics, the acquisition control parameters of the motion camera during the data acquisition process are optimized.

10. A camera motion data set acquisition system, characterized in that: The system comprises: A trend query module, used to obtain initial parameters corresponding to the motion camera, determine the shooting range trajectory corresponding to the motion camera based on the initial parameters, analyze the motion mode corresponding to the shooting range trajectory, and query the motion trajectory trend corresponding to the motion mode; A speed value calculation module is used to perform trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identify trajectory feature parameters in the trajectory feature sequence, and calculate the motion speed value corresponding to the trajectory feature parameter; an element extraction module, configured to analyze a shooting scene corresponding to the motion camera based on the motion speed value, perform scene evaluation on the shooting scene to obtain a scene evaluation value, and extract key scene elements in the shooting scene based on the scene evaluation value; an optimization ratio calculation module, used to determine the acquisition path corresponding to the motion camera based on the key scene elements, detect the path interference item corresponding to the acquisition path, optimize the acquisition control parameters of the motion camera during the data acquisition process based on the path interference item, and calculate the acquisition optimization ratio corresponding to the acquisition control parameters; A scheme generating module is used to generate a data acquisition seed corresponding to the motion camera based on the acquisition optimization ratio, query the detailed acquisition parameters corresponding to the data acquisition seed, identify the acquisition identifier corresponding to the detailed acquisition parameters, and generate a data acquisition scheme corresponding to the motion camera based on the acquisition identifier.

Images