A method and system for collecting camera motion datasets

Abstract

The present invention relates to the field of computer vision and discloses a method and system for collecting camera motion data sets, including: First, obtain the initial parameters of the camera, determine the shooting range trajectory and analyze its motion pattern and trajectory trend, obtain the trajectory feature sequence through trajectory analysis, and calculate the motion speed value. Then, analyze the shooting scene based on this, extract key scene elements after evaluation, and further determine the acquisition path. Detect and optimize the acquisition control parameters based on path interference items, calculate the acquisition optimization ratio, and finally generate data acquisition seeds according to this ratio, query the detailed parameters and acquisition identifiers, so as to generate a data acquisition plan. The present invention can improve the acquisition efficiency of camera motion data.

Description

Technical Field

[0001] The present invention relates to a method and system for collecting a camera motion data set, belonging to the field of computer vision. Background Art

[0002] In the current era of rapid development of digital imaging and intelligent vision technologies, as a key device for obtaining image information, cameras have continuously expanded their application scenarios, covering many fields such as autonomous driving, virtual reality, robot navigation, and film production.

[0003] Currently, the traditional camera motion data collection methods are relatively single. Most rely 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 terms of diversity and complexity, making it difficult to meet the growing technological needs. In addition, some collection methods based on specific scenarios or devices, although they can obtain a certain amount of data under specific conditions, due to scene limitations and device compatibility issues, the collected data lacks generality and extensiveness and cannot effectively support algorithm research and development in a variety of complex application scenarios. Therefore, a method for collecting a camera motion data set is needed to improve the collection efficiency of camera motion data. Summary of the Invention

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

[0005] To achieve the above object, a method for collecting a camera motion data set provided by the present invention includes:

[0006] Obtain the initial parameters corresponding to the moving camera. Based on the initial parameters, determine the shooting range trajectory corresponding to the moving camera, analyze the motion mode corresponding to the shooting range trajectory, and query the motion trajectory trend corresponding to the motion mode;

[0007] Perform trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identify the trajectory feature parameters in the trajectory feature sequence, and calculate the motion speed value corresponding to the trajectory feature parameters;

[0008] Based on the motion speed value, analyze the shooting scene corresponding to the moving camera, perform scene evaluation on the shooting scene to obtain a scene evaluation value, and based on the scene evaluation value, extract the key scene elements in the shooting scene;

[0009] Based on the key scenario elements, determine the acquisition path corresponding to the action camera, detect the path interference items corresponding to the acquisition path, optimize the acquisition control parameters during the data acquisition process of the action camera based on the path interference items, and calculate the acquisition optimization ratio corresponding to the acquisition control parameters;

[0010] Based on the acquisition optimization ratio, generate the data acquisition seeds corresponding to the action camera, query the detailed acquisition parameters corresponding to the data acquisition seeds, and identify the acquisition identifiers corresponding to the detailed acquisition parameters. Based on the acquisition identifiers, generate the data acquisition scheme corresponding to the action camera.

[0011] Optionally, the querying the motion trajectory trend corresponding to the motion mode includes:

[0012] Extract the pattern feature vectors in the motion mode;

[0013] Identify the vector influence factors corresponding to the pattern feature vectors;

[0014] Based on the vector influence factors, determine the motion constraint conditions corresponding to the motion mode;

[0015] Extract the dominant motion variables in the motion constraint conditions;

[0016] Based on the dominant motion variables, query the motion trajectory trend corresponding to the motion mode.

[0017] Optionally, the performing trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence includes:

[0018] Perform discretization processing on the motion trajectory trend to obtain a set of discrete trajectory points;

[0019] Calculate the trajectory change amount between adjacent trajectory points in the set of discrete trajectory points;

[0020] Construct a trajectory change matrix corresponding to the trajectory change amount;

[0021] Extract the core eigenvalues in the trajectory change matrix;

[0022] Based on the core eigenvalues, perform trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence.

[0023] Optionally, the calculating the trajectory change amount between adjacent trajectory points in the set of discrete trajectory points includes:

[0024] Use the following formula to calculate the trajectory change amount between adjacent trajectory points in the set of discrete trajectory points: Where, represents the change in the trajectory between adjacent trajectory points in the set of discrete trajectory points, represents the time node corresponding to the set of discrete trajectory points, represents the quantity indices of the three dimensions of the x, y, and z axes of the spatial dimension, represents at the time node the value of the trajectory point on the k-th dimension, represents at the time node the value of the trajectory point on the k-th dimension, represents the time interval between adjacent time nodes, represents the proportionality coefficient.

[0025] Optionally, calculating the motion speed value corresponding to the trajectory feature parameter includes:

[0026] Calculating the motion speed value corresponding to the trajectory feature parameter using the following formula: where, represents the motion speed value corresponding to the trajectory feature parameter, represents the total number of distance dimension parameters corresponding to the trajectory feature parameter, represents the quantity index corresponding to the distance dimension parameter, represents the trajectory length value corresponding to the n-th distance dimension parameter, represents the weight value corresponding to the n-th distance dimension parameter, represents the total number of resistance dimension parameters corresponding to the trajectory feature parameter, represents the quantity index corresponding to the resistance dimension parameter, represents the quantization value corresponding to the m-th resistance dimension parameter.

[0027] Optionally, analyzing the shooting scene corresponding to the moving camera based on the motion speed value includes:

[0028] Dividing the motion speed value into intervals to obtain a set of speed intervals;

[0029] Identifying the typical scene features of each interval in the set of speed intervals;

[0030] Analyzing the key points of the picture requirements corresponding to the typical scene features;

[0031] Based on the key points of the picture requirements, determining the shooting parameter requirements corresponding to the moving camera;

[0032] Analyzing the shooting scene corresponding to the moving camera based on the shooting parameter requirements.

[0033] Optionally, based on the scene evaluation value, extracting key scene elements in the shooting scene includes:

[0034] Analyzing the scene evaluation metrics corresponding to the scene evaluation value;

[0035] Querying the index distribution corresponding to the scene evaluation metrics;

[0036] Based on the index distribution, determining the scene threshold range corresponding to the scene evaluation metrics;

[0037] Based on the scene threshold range, extracting key scene elements in the shooting scene.

[0038] Optionally, based on the key scene elements, determining the acquisition path corresponding to the action camera includes:

[0039] Identifying the spatial position information of the key scene elements in the shooting scene;

[0040] Extracting the set of position coordinates in the spatial position information;

[0041] Performing regional analysis on the set of position coordinates to obtain the key coordinate region;

[0042] Analyzing the shooting focus corresponding to the key coordinate region;

[0043] Based on the shooting focus, determining the acquisition path corresponding to the action camera.

[0044] Optionally, based on the path interference items, optimizing the acquisition control parameters of the action camera during data acquisition includes:

[0045] Identifying the item type corresponding to the path interference items;

[0046] Based on the item type, collecting the initial control parameters of the action camera in the acquisition path;

[0047] Based on the initial control parameters, determining the performance optimization dimension of the action camera in the control scenario;

[0048] Querying the set of optimization strategies corresponding to the data in the performance optimization dimension;

[0049] Extracting the key parameter features of each sub-strategy data in the set of optimization strategies;

[0050] Based on the key parameter features, optimizing the acquisition control parameters of the action camera during data acquisition.

[0051] To solve the above problems, the present invention also provides an acquisition system based on a camera motion data set, and the system includes:

[0052] A trend query module, configured to obtain initial parameters corresponding to a sports camera, determine a shooting range trajectory corresponding to the sports camera based on the initial parameters, analyze a motion pattern corresponding to the shooting range trajectory, and query a motion trajectory trend corresponding to the motion pattern;

[0053] A speed value calculation module, configured 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 a motion speed value corresponding to the trajectory feature parameters;

[0054] An element extraction module, configured to analyze a shooting scene corresponding to the sports 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;

[0055] An optimization ratio calculation module, configured to determine a collection path corresponding to the sports camera based on the key scene elements, detect path interference items corresponding to the collection path, optimize collection control parameters during data collection of the sports camera based on the path interference items, and calculate a collection optimization ratio corresponding to the collection control parameters;

[0056] A solution generation module, configured to generate a data collection seed corresponding to the sports camera based on the collection optimization ratio, query detailed collection parameters corresponding to the data collection seed, identify a collection identifier corresponding to the detailed collection parameters, and generate a data collection solution corresponding to the sports camera based on the collection identifier.

[0057] Compared with the problems described in the background art, the present invention obtains the initial parameters corresponding to the action camera, which can provide a basic basis for determining the shooting range trajectory subsequently, ensure that the trajectory planning is more accurate and reasonable, help to accurately analyze the motion mode, make the query of the trajectory trend more targeted and effective. The present invention analyzes the motion trajectory trend to obtain a trajectory feature sequence, which can extract key features from the complex motion trajectory, and can more clearly insight into the motion details of the action camera, such as the turning points of the motion, the speed change nodes, etc., helping to identify abnormal situations in the motion mode, and providing strong support for optimizing the data acquisition scheme and improving the shooting effect. Further, based on the motion speed value, the present invention analyzes the shooting scene corresponding to the action camera, can judge whether the moving speed of the camera is appropriate in a specific scene, ensure the stability and clarity of the picture, and can also infer the scene features according to the speed change. For example, a dynamic scene may be captured during rapid movement, helping to optimize the shooting scheme and improve the shooting quality and effect. Further, based on the key scene elements, the present invention determines the acquisition path corresponding to the action camera, 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, avoids ineffective shooting, and the reasonably planned acquisition path helps to present the best perspective of the scene and coherently display the key elements. Finally, based on the acquisition optimization ratio, the present invention generates the data acquisition seeds corresponding to the action camera, which can improve the efficiency of data acquisition. By reasonably planning the acquisition seeds, unnecessary repeated acquisitions or omission of important areas are reduced, making the acquisition process more efficient and orderly. Therefore, a method and system for acquisition based on a camera motion data set provided by an embodiment of the present invention can improve the acquisition efficiency of camera motion data. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 FIG. is a schematic flowchart of a method for acquisition based on a camera motion data set provided by an embodiment of the present invention;

[0059] Figure 2 FIG. is a schematic block diagram of a system for acquisition based on a camera motion data set provided by an embodiment of the present invention.

[0060] The implementation, functional features and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0061] 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.

[0062] The embodiments of the present application provide a method for collecting a camera motion data set. The execution subject of the method for collecting a camera motion data set includes, but is not limited to, at least one of electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiments of the present application. In other words, the method for collecting a 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.

[0063] Embodiment 1:

[0064] Referring to Figure 1 As shown, it is a schematic flowchart 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:

[0065] S1. Obtain the initial parameters corresponding to the motion camera. Based on the initial parameters, determine the shooting range trajectory corresponding to the motion camera, analyze the motion mode corresponding to the shooting range trajectory, and query the motion trajectory trend corresponding to the motion mode.

[0066] By obtaining the initial parameters corresponding to the motion camera in the present invention, it can provide a basic basis for subsequent determination of the shooting range trajectory, ensure more accurate and reasonable trajectory planning, help accurately analyze the motion mode, and make the query of the motion trajectory trend more targeted and effective.

[0067] Among them, the motion camera refers to a camera device that can move and is used to collect image or video data. It has the ability to work in a dynamic environment. For example, it is used to monitor the surrounding road conditions on an autonomous vehicle, assist in navigation on a robot, or capture dynamic pictures in film and television shooting. The motion camera is usually equipped with various sensors, such as an accelerometer, a gyroscope, etc.; the initial parameters refer to a series of set or measured values that need to be obtained before starting to collect motion camera data. These parameters include, but are not limited to, the installation position coordinates of the camera, the initial orientation angle, the lens focal length, the frame rate, the resolution, etc. The installation position coordinates determine the starting point of the camera in space, the initial orientation angle clarifies the initial shooting direction of the camera, the lens focal length affects the size and range of the imaging, the frame rate determines the number of images collected per unit time, and the resolution is related to the clarity of the collected images or videos. Optionally, the obtaining of the initial parameters corresponding to the motion camera can be realized by the built-in sensors of the camera, such as tools like an accelerometer, a gyroscope, etc.

[0068] Furthermore, based on the initial parameters, the present invention determines the shooting range trajectory corresponding to the action camera and analyzes the motion mode corresponding to the shooting range trajectory, which can accurately plan the shooting content, ensure that the captured images or videos meet the expected requirements, and provide effective materials for subsequent data processing.

[0069] Among them, the shooting range trajectory refers to the spatial range covered by the lens of the action camera during shooting and the moving path of this coverage range over time. It comprehensively considers factors such as 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 scenario of an autonomous vehicle, when the action camera is installed at the front of the vehicle, its shooting range trajectory can start from an initial rectangular area in front of the vehicle head and gradually extend forward and swing left and right by a certain angle as the vehicle travels, forming a dynamic three-dimensional space trajectory; the motion mode refers to the motion laws and ways followed by the action camera during shooting, which reflects the characteristics and features of the camera's motion, including translational motion (such as uniform linear translation, variable-speed curve translation, etc.), rotational motion (such as uniform rotation around a fixed axis, non-uniform rotation, etc.), compound motion (complex motion with translation and rotation occurring simultaneously), etc. Optionally, the determination of the shooting range trajectory corresponding to the action camera can be achieved through feature detection algorithms, such as SIFT, SURF, etc.; the analysis of the motion mode corresponding to the shooting range trajectory can be achieved through machine learning algorithms, such as support vector machines, decision trees, etc.

[0070] Furthermore, by querying the motion trajectory trend corresponding to the motion mode, the present invention helps to anticipate the motion direction of the camera in advance, provides a forward-looking decision-making basis for related applications (such as path planning in autonomous driving, camera scheduling in film shooting), and makes the operation more accurate and efficient.

[0071] Among them, the motion trajectory trend describes the possible development direction and changes of the future motion trajectory of the camera in the current motion mode. By analyzing information such as the dominant motion variables, for example, whether it is continuous accelerating linear motion, decelerating turning motion, or maintaining uniform circular motion, etc.

[0072] As an embodiment of the present invention, the querying of the motion trajectory trend corresponding to the motion mode includes: extracting the mode feature vector in the motion mode; identifying the vector influencing factor corresponding to the mode feature vector; determining the motion constraint condition corresponding to the motion mode based on the vector influencing factor; extracting the dominant motion variable in the motion constraint condition; and querying the motion trajectory trend corresponding to the motion mode based on the dominant motion variable.

[0073] Among them, the pattern feature vector refers to a mathematical vector that can quantitatively describe the characteristics of a motion pattern, which contains various feature information related to the motion pattern, such as the speed, acceleration, direction change frequency, motion amplitude, etc. of the motion; the vector influence factor refers to the factor that plays a key influencing role in determining the motion pattern and subsequent analysis in the pattern feature vector. For example, in the feature vector describing the camera rotation motion pattern, the value of the rotation speed dimension and its change rate over time are important vector influence factors; the motion constraint conditions refer to a series of conditions that restrict the camera motion determined based on the vector influence factors, and these conditions are derived from the information contained in the vector influence factors, including but not limited to the spatial range limit of the motion, such as the camera can only rotate within a specific angle range; the speed limit, that is, the camera motion speed cannot exceed a certain threshold; and the time limit, such as completing a specific motion action within a specific time period, etc.; the dominant motion variable refers to the variable that plays a major determining role in the trend of the motion trajectory extracted from the motion constraint conditions, and it is the core element among many motion constraint conditions. For example, in the linear translation motion pattern of the camera, the translation direction and speed are the dominant motion variables.

[0074] Furthermore, the extraction of the pattern feature vector in the motion pattern can be achieved through the principal component analysis algorithm. For example, these data can be dimensionally reduced to obtain the pattern feature vector while retaining the main information; the identification of the vector influence factor corresponding to the pattern feature vector can be achieved through the neural network method. For example, using a multi-layer perceptron (MLP), taking the pattern feature vector and possible influencing factors as the input layer, through the non-linear transformation of the hidden layer, the weights related to the influencing factors are obtained in the output layer, and these weights can be used as the vector influence factors; the determination of the motion constraint conditions corresponding to the motion pattern can be achieved through the CSP algorithm. For example, constructing a CSP model with the various variables and constraint conditions of the motion pattern, and finding the solution that satisfies all the constraint conditions through the search algorithm to determine the motion constraint conditions; the extraction of the dominant motion variable in the motion constraint conditions can be achieved through the particle swarm optimization algorithm. For example, the motion constraint conditions can be transformed into a fitness function, and the variable combination that makes the fitness optimal (that is, the most dominant) can be searched through the PSO algorithm to extract the dominant motion variable; the query of the motion trajectory trend corresponding to the motion pattern can be achieved through the clustering analysis method. For example, clustering similar motion trajectories, and then analyzing the trend of the central trajectory of each cluster to determine the overall motion trajectory trend.

[0075] S2. Conduct trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identify the trajectory feature parameters in the trajectory feature sequence, and calculate the motion speed value corresponding to the trajectory feature parameters.

[0076] Through trajectory analysis of the motion trajectory trend, the present invention obtains a trajectory feature sequence, which can extract key features from complex motion trajectories, enabling a clearer insight into the motion details of the action camera, such as turning points of motion, speed change nodes, etc., helping to identify anomalies in the motion pattern and providing strong support for optimizing the data acquisition scheme and improving the shooting effect.

[0077] Among them, the trajectory feature sequence refers to a series of ordered combinations of eigenvalue obtained by analyzing the motion trajectory trend based on the core eigenvalue. It is a feature description of the entire motion trajectory, containing key information of the motion trajectory, such as features like motion mode, frequency, amplitude, etc.

[0078] 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 set of discrete trajectory points; calculating the trajectory change amount between adjacent trajectory points in the set of discrete trajectory points; constructing a trajectory change matrix corresponding to the trajectory change amount; extracting the core eigenvalue from the trajectory change matrix; and based on the core eigenvalue, performing trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence.

[0079] Among them, the set of discrete trajectory points refers to a set of a series of discrete points obtained by dividing the continuous motion trajectory trend at a certain time interval or spatial interval. For example, recording the position of the action camera every 0.1 seconds, and these position points constitute the set of discrete trajectory points; the trajectory change amount refers to the change values in terms of position, speed, acceleration, etc. between adjacent trajectory points in the set of discrete trajectory points. For example, the displacement difference between two adjacent points represents the change amount of position, and the speed difference is the change amount of speed; the trajectory change matrix refers to a matrix constructed with the trajectory change amount as elements. For example, the first row of the matrix can represent the displacement change amount between adjacent points, and the second row represents the speed change amount, etc.; the core eigenvalue refers to the representative and important values extracted from the trajectory change matrix, and these values can prominently reflect the main features of the motion trajectory, such as the periodicity of motion, mutation points, etc.

[0080] Furthermore, the discretization of the motion trajectory trend can be achieved by the equidistant division method. For example, the range of the entire motion trajectory is divided into several small rectangular regions at fixed distance intervals, and the vertices or center points of each region can be used as discrete trajectory points. The calculation of the trajectory change amount between adjacent trajectory points in the set of discrete trajectory points can be achieved by the numerical differentiation method. For example, after calculating the velocity and acceleration of each point using the central difference method, the velocity change amount and acceleration change amount between adjacent points are further calculated, which together with the position change amount form the trajectory change amount. The construction of the trajectory change matrix corresponding to the trajectory change amount can be achieved by the matrix filling method. For example, the position change amount between the i-th and the (i + 1)-th trajectory points is filled in the first column of the i-th row, the velocity change amount is filled in the second column, and the acceleration change amount is filled in the third column, finally obtaining the trajectory change matrix. The extraction of the core eigenvalue in the trajectory change matrix can be achieved by the eigenvalue decomposition method. For example, in Python, the linalg.eigh function of the SciPy library is used 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 achieved by the machine learning classification method, such as algorithms like support vector machines and decision trees.

[0081] As an embodiment of the present invention, the calculation of the trajectory change amount between adjacent trajectory points in the set of discrete trajectory points includes:

[0082] The trajectory change amount between adjacent trajectory points in the set of discrete trajectory points is calculated using the following formula: where, represents the trajectory change amount between adjacent trajectory points in the set of discrete trajectory points, represents the time node corresponding to the set of discrete trajectory points, represents the number indices of the three dimensions of the x, y, and z axes in the spatial dimension, represents at the time node the value of the trajectory point in the k-th dimension, represents at the time node the value of the trajectory point in the k-th dimension, represents the time interval between adjacent time nodes, represents the proportionality coefficient.

[0083] Furthermore, by identifying the trajectory feature parameters in the trajectory feature sequence, the present invention can help accurately grasp the key information of the motion trajectory of the moving camera, such as special nodes of motion, change rules, etc. These feature parameters also provide core data support for subsequent calculation of the motion speed value, analysis of the shooting scene, etc., helping to understand the motion situation more deeply and laying a solid foundation for optimizing the data acquisition scheme.

[0084] Among them, the trajectory feature parameters refer to the key parameters identified from the trajectory feature sequence that can describe the characteristics of the motion trajectory. These parameters contain the characteristic information of the motion trajectory in aspects such as position, velocity, acceleration, and direction. For example, the position coordinates of specific points in the trajectory, the velocity peak value, the change rate of acceleration, the change angle of the motion direction, etc. all belong to the trajectory feature parameters, which can reflect the shape of the motion trajectory, the speed change of the motion, and the dynamic change trend of the trajectory. Optionally, the identification of the trajectory feature parameters in the trajectory feature sequence can be achieved through parameter extraction tools, such as tools like MATLAB, Python, etc.

[0085] Furthermore, by calculating the motion speed value corresponding to the trajectory feature parameters, the present invention can intuitively reflect the motion speed of the moving camera at a specific trajectory stage, clearly understand its dynamic changes, provide a key basis for analyzing the shooting scene, and help judge the rationality of the camera's motion in different scenarios.

[0086] Among them, the motion speed value refers to the numerical value that reflects the speed magnitude of the moving camera under a specific motion trajectory. It comprehensively considers various factors such as the distance dimension parameter and the resistance dimension parameter, and is a quantified motion speed characterization quantity used to describe the speed of the camera during the motion process.

[0087] As an embodiment of the present invention, the calculation of the motion speed value corresponding to the trajectory feature parameters includes:

[0088] Calculating the motion speed value corresponding to the trajectory feature parameters using the following formula: Among them, represents the motion speed value corresponding to the trajectory feature parameters, represents the total number of distance dimension parameters corresponding to the trajectory feature parameters, represents the quantity index corresponding to the distance dimension parameter, represents the th trajectory length value corresponding to the th distance dimension parameter, represents the weight value corresponding to the th distance dimension parameter, represents the total number of resistance dimension parameters corresponding to the trajectory feature parameters, represents the quantity index corresponding to the resistance dimension parameter, represents the quantization value corresponding to the

[0089] Specifically, the distance dimension parameter refers to a type of parameter related to distance among the trajectory feature parameters. For example, in a three-dimensional space, it corresponds to the distance-related quantities in the x, y, and z directions respectively, and can also include distance measurement parameters such as the arc length of a curved trajectory. The trajectory length value refers to, for the i-th distance dimension parameter, the length measurement value of the movement trajectory in this dimension. For example, in the x-axis direction, the straight-line distance or the curve length that the camera moves from one point to another is the trajectory length value corresponding to the distance dimension parameter of the x-axis. The weight value refers to a coefficient corresponding to the i-th distance dimension parameter, which is used to measure the relative importance of this distance dimension parameter when calculating the movement speed value, and its value range is generally between 0 and 1. If a certain distance dimension parameter has a greater impact on the movement speed, a higher weight value can be assigned to it; conversely, if the impact is smaller, the weight value is lower. The resistance dimension parameter refers to a type of parameter related to the resistance suffered during the movement among the trajectory feature parameters. These parameters describe various resistance factors that may affect the movement speed, such as air resistance, ground friction, and the resistance generated by obstacles encountered during the movement. The quantization value refers to, for the j-th resistance dimension parameter, the value obtained by quantifying this resistance factor. For example, for air resistance, a specific value may be calculated based on factors such as the density of the air, the windward area of the moving object, and the speed to represent the magnitude of the air resistance, and this value is the quantization value corresponding to the resistance dimension parameter of air resistance.

[0090] S3. Based on the movement speed value, analyze the shooting scene corresponding to the moving camera, conduct a scene evaluation on the shooting scene to obtain a scene evaluation value, and based on the scene evaluation value, extract the key scene elements in the shooting scene.

[0091] Based on the movement speed value, the present invention analyzes the shooting scene corresponding to the moving camera, can judge whether the moving speed of the camera is appropriate in a specific scene to ensure the stability and clarity of the picture, and can also infer the scene characteristics based on the speed change. For example, a dynamic scene may be captured during rapid movement, which helps to optimize the shooting plan and improve the shooting quality and effect.

[0092] Among them, the shooting scene refers to the specific environment and conditions suitable for the moving camera to shoot based on the shooting parameter requirements, including the actual physical environment (such as indoor, outdoor, different terrains, etc.) and the setting state of the camera. For example, arranging the lights indoors, adjusting the position and angle of the camera according to the shooting parameter requirements constitutes a specific shooting scene.

[0093] As an embodiment of the present invention, analyzing the shooting scene corresponding to the action camera based on the motion speed value includes: dividing the motion speed value into intervals to obtain a set of speed intervals; identifying the typical scene features of each interval in the set of speed intervals; analyzing the key points of the picture requirements corresponding to the typical scene features; determining the shooting parameter requirements corresponding to the action camera based on the key points of the picture requirements; and analyzing the shooting scene corresponding to the action camera based on the shooting parameter requirements.

[0094] Among them, the set of speed intervals refers to the 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 m / s), a medium-speed interval (5 - 15 m / s), a high-speed interval (15 - 30 m / s), etc. These different speed intervals together constitute the set of speed intervals; the typical scene features refer to the common scene characteristics corresponding to each interval in the set of speed intervals. For example, in the low-speed interval, the typical scene features can be shooting static objects or slowly moving objects, such as still life shooting indoors, slow walking of people, etc.; the medium-speed interval can correspond to shooting general dynamic scenes, such as vehicles moving normally on urban streets, etc.; the high-speed interval can correspond to shooting fast-moving scenes, such as fast running in sports events, etc.; the key points of the picture requirements refer to the aspects that need to be focused on when shooting the picture based on the typical scene features. For example, for the static scene corresponding to the low-speed interval, the key points of the picture requirements can be high resolution to clearly present the details of the object; for the fast-moving scene corresponding to the high-speed interval, the key points of the picture requirements can be high frame rate to avoid blurry pictures and ensure the coherence of moving objects; the shooting parameter requirements refer to the specific parameter setting requirements of the action camera determined according to the key points of the picture requirements, covering parameters such as aperture size, shutter speed, sensitivity (ISO), focal length, etc. For example, when the key point of the picture requirement is high frame rate, the shooting parameter requirements can be to set a higher shutter speed and appropriate sensitivity to ensure the quality and stability of the picture during fast shooting.

[0095] Furthermore, the interval division of the motion speed value can be achieved by the equal-distance division method. For example, if the range of the motion speed value is 0 - 50 m / s and the set interval is 10 m / s, 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 recognition of the typical scene features in each interval of the speed interval set can be achieved by machine learning classification methods, such as classification algorithms like support vector machines and random forests. The analysis of the key points of the picture requirements corresponding to the typical scene features can be achieved by a decision tree model. For example, by training a decision tree model, when new typical scene features are input, the corresponding key points of the picture requirements can be quickly determined. The determination of the shooting parameter requirements corresponding to the action camera can be achieved by the mapping relationship establishment method. For example, by establishing a mapping table between the key points of the picture requirements and the shooting parameters, the corresponding shooting parameter requirements can be determined according to the key points of the picture requirements obtained previously. The analysis of the shooting scene corresponding to the action camera can be achieved by project management tools, such as tools like Jira. Taking the results of each link as task inputs, by setting task dependencies and workflows, the shooting scene corresponding to the action camera can be finally summarized.

[0096] Through the scene evaluation of the shooting scene, the present invention obtains a scene evaluation value, which can systematically measure the quality of the shooting scene, clarify whether the current scene fits the shooting target, and through the evaluation value, problems in the scene, such as insufficient light and cluttered background, can also be quickly discovered, and then targeted adjustments and optimizations can be made.

[0097] Among them, the scene evaluation value refers to a quantified numerical value 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 criteria and weights. A higher scene evaluation value means that the scene performs well in aspects such as sufficient light, favorable space for framing, simple background, and harmonious color, which is more conducive to obtaining high-quality shooting pictures; on the contrary, it indicates that there are some areas in the scene that need to be improved or adjusted to meet the shooting requirements. Optionally, the scene evaluation of the shooting scene can be achieved by scene evaluation methods, such as the analytic hierarchy process, fuzzy comprehensive evaluation method, etc.

[0098] Furthermore, based on the scene evaluation value, the present invention extracts the key scene elements in the shooting scene, which can focus on the core content, can be optimized and highlighted in a targeted manner, helps with the effective organization and analysis of data, and enables it to automatically adjust shooting parameters and strategies according to the key elements to better adapt to the scene requirements.

[0099] 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 such as light intensity and color contrast, there may be some elements with unique visual effects or closely related to the shooting theme, such as prominent main objects, guiding lines, special light and shadow effects, etc.

[0100] As an embodiment of the present invention, extracting the key scene elements in the shooting scene based on the scene evaluation value includes: analyzing the scene evaluation indicators corresponding to the scene evaluation value; querying the index distribution of the scene evaluation indicators; determining the scene threshold range corresponding to the scene evaluation indicators based on the index distribution; and extracting the key scene elements in the shooting scene based on the scene threshold range.

[0101] Among them, the scene evaluation indicators refer to 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 quantitatively evaluating the shooting scene; the index distribution refers to the numerical distribution state of each scene evaluation indicator in the entire shooting scene, which describes the change of each indicator in different regions or conditions. For example, for the light intensity indicator, its distribution can show which areas in the shooting scene are brighter and which are darker; for the color contrast indicator, the distribution can reflect the high and low distribution of the contrast between different color regions; the scene threshold range refers to a numerical interval determined according to the index distribution and shooting requirements. When the value of an indicator falls within this range, it indicates that the scene conditions corresponding to the indicator meet certain shooting requirements or standards. For example, for a scene of taking a close-up of a person, the scene threshold range of light intensity may be set within a specific interval to ensure that the light on the person's face is clear and there are no shadows.

[0102] Furthermore, the analysis of the scene evaluation indicators corresponding to the scene evaluation value can be achieved through feature selection algorithms, such as: algorithms like chi-square test, mutual information method, etc.; the query of the index distribution of the scene evaluation indicators can be achieved through data analysis tools, such as: tools like LabVIEW, MATLAB, etc.; the determination of the scene threshold range corresponding to the scene evaluation indicators can be achieved through the threshold determination method of machine learning, such as: using the SVM algorithm in the scikit-learn library, and determining the appropriate threshold ranges of indicators such as light uniformity and background complexity after learning a large amount of shooting scene data; the extraction of the key scene elements in the shooting scene can be achieved through deep learning frameworks, such as: frameworks like TensorFlow, PyTorch, etc.

[0103] S4. Based on the key scene elements, determine the acquisition path corresponding to the action camera, detect the path interference items corresponding to the acquisition path, optimize the acquisition control parameters during the data acquisition process of the action camera based on the path interference items, and calculate the acquisition optimization ratio corresponding to the acquisition control parameters.

[0104] Based on the key scene elements, the present invention determines the acquisition path corresponding to the action camera, 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, avoids ineffective shooting, and the reasonably planned acquisition path helps to present the best perspective of the scene and coherently display the key elements.

[0105] Among them, the acquisition path refers to the moving route planned by the action camera in the shooting scene in order to comprehensively and effectively capture the key scene elements. For example, when shooting a large building, the acquisition path may move in a circular pattern around the building to capture the key parts of the building from different angles, such as the unique building appearance, delicate decorative details, etc.

[0106] As an embodiment of the present invention, the determining the acquisition path corresponding to the action camera based on the key scene elements includes: identifying the spatial position information of the key scene elements in the shooting scene; extracting the set of position coordinates in the spatial position information; performing regional analysis on the set of position coordinates to obtain the key coordinate region; analyzing the shooting focus corresponding to the key coordinate region; and determining the acquisition path corresponding to the action camera based on the shooting focus.

[0107] Among them, the spatial position information refers to the specific orientation description of key scene elements 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 after quantifying the spatial position information of key scene elements 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), and multiple such coordinate values form the position coordinate set; the key coordinate region refers to the specific region divided after performing regional analysis on the position coordinate set. For example, when shooting a sports event, the area where athletes gather and move is divided into a key coordinate region after analyzing the position coordinate sets of relevant key scene elements (athletes, balls, etc.); the shooting focus refers to the key attention direction determined according to the characteristics of the key coordinate region and the expected shooting goal. Different key coordinate regions may have different characteristics. For example, some regions have rich elements and large dynamic changes, while some regions have bright colors and high contrast. The shooting focus can include highlighting the main element, capturing dynamic moments, showing color levels, etc.

[0108] 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 coordinate conversion algorithms, 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 clustering algorithms, such as: K-Means, DBSCAN and other tools; the analysis of the shooting focus corresponding to the key coordinate region can be achieved through a rule-based reasoning method, such as: the judgment and reasoning of these rules can be achieved by writing a Python program to obtain the shooting focus corresponding to each key coordinate region; the determination of the acquisition path corresponding to the motion camera can be achieved through the Dijkstra algorithm, such as: setting each key coordinate region as a node in the graph, and using the Dijkstra algorithm to start from the starting node and calculate the shortest paths to all other nodes, so as to obtain the acquisition path of the motion camera.

[0109] By detecting the path interference items corresponding to the acquisition path, the present invention can discover in advance the factors that can hinder the smooth movement of the motion camera, such as obstacles, signal interference sources, etc., so as to adjust the path planning in a timely manner, avoid shooting interruption or problems such as shaking and blurring of the picture, and by effectively avoiding the interference items, it can ensure the smooth progress of the acquisition work, improve the shooting efficiency, and provide a strong guarantee for obtaining high-quality video materials.

[0110] Among them, the path interference items refer to various factors that can have an adverse impact on the shooting process of the action camera when it operates according to the established acquisition path. These factors cover a wide range, including physical obstacles, such as pedestrians and tree 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 and wireless signal transmission devices, which can interfere with the signal transmission and control of the camera, resulting in problems such as abnormal shooting parameters and data loss; in addition, unstable environmental factors, such as strong winds and uneven ground, also belong to the path interference items. Optionally, detecting the path interference items corresponding to the acquisition path can be achieved through a sensor-based detection method. For example, lidar emits laser beams around and measures the time when the laser reflects back to the sensor to construct the three-dimensional point cloud data of the surrounding environment, and by analyzing the point cloud data, interference items such as obstacles on the acquisition path can be detected.

[0111] Furthermore, based on the path interference items, the present invention optimizes the acquisition control parameters of the action camera during the data acquisition process, which can significantly improve the shooting effect. Through targeted adjustment, the impact of interference on the image quality can be effectively avoided. For example, adjusting the exposure parameters under light interference to ensure a clear picture, preventing data loss or abnormality caused by interference, and enabling the acquisition work to proceed smoothly to obtain high-quality image data.

[0112] Among them, the acquisition control parameters refer to various parameters that can be adjusted by the action 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.

[0113] As an embodiment of the present invention, optimizing the acquisition control parameters of the action camera during the data acquisition process based on the path interference items includes: identifying the item type corresponding to the path interference item; based on the item type, acquiring the initial control parameters of the action camera in the acquisition path; based on the initial control parameters, determining the performance optimization dimension of the action camera in the control scenario; querying the optimization strategy set corresponding to the data in the performance optimization dimension; extracting the key parameter features of each sub-strategy data in the optimization strategy set; and optimizing the acquisition control parameters of the action camera during the data acquisition process based on the key parameter features.

[0114] 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. For example, the path interference items are classified into categories such as entity obstacle interference, light interference, signal interference, etc.; the initial control parameters refer to the parameter setting values of the motion camera before optimizing for the current path interference item during the acquisition path. These parameters cover multiple 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 determined according to the initial control parameters and the path interference item. It is a classification and detailed description of the camera performance. For example, when facing light interference, the performance optimization dimension can be the brightness uniformity and color reproduction of the image, etc.; when facing entity obstacle interference, it can be dimensions such as the focusing speed and viewing angle range of the camera; the optimization strategy set refers to a set of a series of feasible optimization methods formulated 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 strategies, local brightness compensation strategies, etc.; the key parameter features refer to the parameter characteristics extracted from the data of each sub-strategy in the optimization strategy set, which are representative and decisive. These features can reflect the core points of each sub-strategy. For example, the parameter features such as the adjustment range of the shutter speed and the change amplitude of the aperture value in a certain optimization strategy.

[0115] Further, the identification of the project type corresponding to the path interference item can be achieved through a rule-based classification method. For example, if the interference item is a physical entity and has a large volume, it can be classified as an "entity 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 action camera in the acquisition path can be achieved through camera API calls. For example, by sending a GET request to the API of the GoPro camera, obtaining the status information of the camera, and extracting the camera settings information from the returned JSON data as the initial control parameters. The determination of the performance optimization dimension of the action camera in the control scenario can be achieved through virtual scene simulation tools, such as tools like Unity and Unreal Engine. The query of the optimization strategy set corresponding to the data in the performance optimization dimension can be achieved through a database query system. For example, a database containing various performance optimization dimensions and their corresponding optimization strategies is established, and by inputting the performance optimization dimension, the corresponding optimization strategy set is queried from the database. The extraction of the key parameter features of each sub-strategy data in the optimization strategy set can be achieved through feature extraction algorithms. For example, for the automatic exposure adjustment algorithm, the key parameter features may include the exposure time adjustment step size, brightness threshold, etc.; for the manual exposure curve setting, the key parameter features may be the exposure values in different brightness intervals. The optimization of the acquisition control parameters of the action camera during the data acquisition process can be achieved through a rule-based parameter adjustment algorithm. For example, if it is determined that the current ambient light is dim and the exposure time adjustment step size of the automatic exposure adjustment algorithm is 0.1 seconds, the exposure time of the action camera is adjusted from 1 / 60 seconds to 1 / 30 seconds, and at the same time, the ISO value is adjusted to 400 according to the brightness threshold, so as to optimize the acquisition control parameters.

[0116] S5. Based on the acquisition optimization ratio, generate the data acquisition seeds corresponding to the action camera, query the detailed acquisition parameters corresponding to the data acquisition seeds, and identify the acquisition identifiers corresponding to the detailed acquisition parameters. Based on the acquisition identifiers, generate the data acquisition scheme corresponding to the action camera.

[0117] Based on the acquisition optimization ratio, the present invention generates the data acquisition seeds corresponding to the action camera, which can improve the efficiency of data acquisition. By reasonably planning the acquisition seeds, unnecessary repeated acquisitions or omission of important areas are reduced, making the acquisition process more efficient and orderly.

[0118] Among them, the data collection seed refers to an initial element or data set that provides basic guidance and key information for the data collection of a sports camera. It is generated based on factors such as the collection optimization ratio and contains key information such as specific collection locations, time points, and shooting parameter preferences. It is like the "guidance seed" for data collection work, which can guide the sports camera to perform data collection in a more optimized and targeted manner, laying a foundation for obtaining high-quality data that meets specific requirements. Optionally, the generation of the data collection seed corresponding to the sports camera can be achieved through a seed generation algorithm, such as genetic algorithm, simulated annealing algorithm, etc.

[0119] Furthermore, by querying the detailed collection parameters corresponding to the data collection seed and identifying the collection identifiers corresponding to the detailed collection parameters, the present invention can significantly improve the standardization and efficiency of data collection, enable the sports camera to accurately execute the collection task, avoid collection deviations caused by unclear parameters, and lay a solid foundation for the effective utilization of data.

[0120] Among them, the detailed collection parameters refer to a set of specific and detailed parameter sets set for the data collection work of the sports camera. These parameters cover all key aspects of camera shooting and are used to precisely guide how the camera conducts data collection. For example, it includes the resolution settings of shooting, such as common 1080p, 4K, etc. Different resolutions determine the clarity and detail richness of images or videos; the 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 collection identifier refers to a specific symbol or code set to distinguish and identify different collection tasks or data sets. It can be a number, letter, string, or combination form. By assigning a unique collection identifier to each collection task, different sources and purposes of data can be clearly distinguished. For example, in a complex environmental monitoring project, different collection identifiers are set for collection tasks in different regions and at different times. Optionally, the query of the detailed collection parameters corresponding to the data collection seed can be achieved through querying a relational database, such as MySQL, Oracle, etc.; the identification of the collection identifier corresponding to the detailed collection parameters can be achieved through the hash function method, such as using a hash function to process the detailed collection parameters to generate a unique hash value as the collection identifier.

[0121] Based on the collection identifier, the present invention generates a data collection scheme corresponding to the sports camera, which can greatly improve the pertinence and systematicness of the collection work. The collection identifier, as a key index, can accurately match relevant information to ensure that the collection scheme fully fits specific scenarios and requirements, enabling the data collection process to be more standardized and orderly, and improving the quality and usability of the finally collected data.

[0122] Among them, the data acquisition scheme refers to a comprehensive plan carefully formulated for the action camera, aiming to ensure efficient and accurate acquisition of the required data. It details various key elements during the data acquisition process of the action camera, including but not limited to: clarifying the goals of the acquisition task, such as for environmental monitoring, behavior analysis, or other specific purposes; defining the acquisition scope, such as a specific area, time period, or specific type of scene, etc.; precisely setting various parameters of the camera, such as resolution, frame rate, sensitivity, white balance, etc., to meet different shooting requirements; planning the acquisition path to guide the movement trajectory of the camera at different spatial positions; setting the triggering conditions for data acquisition, such as the occurrence of a specific event, reaching a certain time point, or meeting specific environmental parameters, etc.; and also covering the storage method, data format, and preliminary data sorting and screening strategies after data acquisition. Optionally, generating the data acquisition scheme corresponding to the action camera can be achieved through scheme generation tools, such as tools like Trello, Asana, etc.

[0123] Compared with the problems described in the background art, the present invention obtains the initial parameters corresponding to the action camera, which can provide a basic basis for subsequent determination of the shooting range trajectory, ensure more accurate and reasonable trajectory planning, help analyze the motion pattern precisely, and make the query of the trajectory trend more targeted and effective. By performing trajectory analysis on the motion trajectory trend, the present invention obtains a trajectory feature sequence, which can extract key features from complex motion trajectories, and can more clearly understand the motion details of the action camera, such as the turning points of motion, speed change nodes, etc., helping to identify abnormal situations in the motion pattern and providing strong support for optimizing the data acquisition scheme and improving the shooting effect. Further, based on the motion speed value, the present invention analyzes the shooting scene corresponding to the action camera, can judge whether the camera moving speed is appropriate in a specific scene, ensure the stability and clarity of the picture, and can also infer the scene features based on the speed change. For example, a dynamic scene may be captured during fast movement, helping to optimize the shooting scheme and improve the shooting quality and effect. Further, based on the key scene elements, the present invention determines the acquisition path corresponding to the action camera, which can greatly improve the pertinence and efficiency of shooting. It ensures that the camera precisely captures the elements crucial for the shooting effect, avoids ineffective shooting, and the reasonably planned acquisition path helps to present the best perspective of the scene and coherently display the key elements. Finally, based on the acquisition optimization ratio, the present invention generates the data acquisition seeds corresponding to the action camera, which can improve the efficiency of data acquisition. By reasonably planning the acquisition seeds, unnecessary repeated acquisitions or omission of important areas are reduced, making the acquisition process more efficient and orderly. Therefore, a method and system for acquisition based on a camera motion data set provided by an embodiment of the present invention can improve the acquisition efficiency of camera motion data.

[0124] Embodiment 2:

[0125] As shown Figure 2 in the figure, it is a functional module diagram of a collection system of the present invention based on a camera motion data set.

[0126] The collection system 200 of the present invention based on a camera motion data set can be installed in an electronic device. According to the functions achieved, the collection system based on a camera motion data set may 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 modules of the present invention may also be referred to as units, which refer to a series of computer program segments that can be executed by a processor of an electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

[0127] In an embodiment of the present invention, the functions of each module / unit are as follows:

[0128] The trend query module 201 is used to obtain the 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;

[0129] The speed value calculation module 202 is used to perform trajectory analysis on the motion trajectory trend to obtain a trajectory feature sequence, identify the trajectory feature parameters in the trajectory feature sequence, and calculate the motion speed value corresponding to the trajectory feature parameters;

[0130] 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;

[0131] 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 items corresponding to the acquisition path, optimize the acquisition control parameters during the data acquisition process of the motion camera based on the path interference items, and calculate the acquisition optimization ratio corresponding to the acquisition control parameters;

[0132] 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, 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.

[0133] Specifically, each module in the acquisition system 200 based on a camera motion data set in the embodiments of the present invention adopts the same technical means as the Figure 1 a method for acquiring a camera motion data set described therein, and can produce the same technical effects, which will not be elaborated here.

[0134] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms.

[0135] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions 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; Analyzing the shooting scene corresponding to the motion camera based on the motion speed value, wherein 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, analyzing the shooting scene corresponding to the motion camera, wherein the shooting scene refers to a specific environment and conditions suitable for shooting with a motion camera constructed based on the shooting parameter requirements, performing scene evaluation on the shooting scene to obtain a scene evaluation value, wherein the scene evaluation value refers to a quantitative value used to comprehensively reflect the suitability of the shooting scene for achieving the expected shooting effect, which is obtained by scoring and calculating multiple key elements in the shooting scene, including: light conditions, spatial layout, background complexity, and color matching according to specific evaluation criteria and weights, 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: Among them, ΔGb t ' represents the trajectory change between adjacent trajectory points in the discrete trajectory point set, t' represents the time node corresponding to the discrete trajectory point set, k represents the number index of the three dimensions of the x, y, and z axes of the spatial dimension, X t',k represents the value of the trajectory point in the kth dimension at time node t', X t'+1,k It represents the value of the trajectory point in the kth dimension at the time node t'+1, ΔT represents the time interval between adjacent time nodes, and ω represents the proportional coefficient.

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: Wherein, V represents the motion speed value corresponding to the trajectory characteristic parameter, n represents the total number of distance dimension parameters corresponding to the trajectory characteristic parameter, i represents the number index corresponding to the distance dimension parameter, C i represents the trajectory length value corresponding to the i-th distance dimension parameter, T i represents the weight value corresponding to the i-th distance dimension parameter, m represents the total number of resistance dimension parameters corresponding to the trajectory feature parameters, j represents the number index corresponding to the resistance dimension parameter, S j Represents the quantitative value corresponding to the j-th resistance dimension parameter.

6. 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.

7. 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.

8. 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.

9. 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 is used to analyze the shooting scene corresponding to the motion camera based on the motion speed value, wherein 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, analyzing the shooting scene corresponding to the motion camera, wherein the shooting scene refers to a specific environment and conditions suitable for shooting with a motion camera constructed based on the shooting parameter requirements, performing scene evaluation on the shooting scene to obtain a scene evaluation value, wherein the scene evaluation value refers to a quantitative value used to comprehensively reflect the suitability of the shooting scene for achieving the expected shooting effect, which is obtained by scoring and calculating multiple key elements in the shooting scene, including: light conditions, spatial layout, background complexity, and color matching according to specific evaluation criteria and weights, and extracting 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