A feedback analysis-based naked-eye 3D imaging evaluation method and system

Abstract

This invention discloses a naked-eye 3D imaging evaluation method and system based on feedback analysis, relating to the field of 3D display technology. A dedicated network is established between the cloud and the 3D terminal to store and convert initial display data into naked-eye 3D content in real time. Based on feedback analysis, on the one hand, an interaction matrix is ​​generated through user interaction behavior to analyze differences in operational characteristics; on the other hand, binocular vision simulation technology is used to analyze the color histogram distribution of the left and right views using a moving pixel window to quantify imaging weights. The system performs feature decomposition on the interaction matrix, establishes content priority groups based on imaging weights, and dynamically adjusts the cloud caching strategy to improve 3D display efficiency. This invention effectively improves user experience and optimizes imaging quality, effectively enhances the screen interaction efficiency of naked-eye 3D display devices, and improves the multi-user display efficiency of 3D display terminals.

Description

Technical Field

[0001] This invention relates to the field of naked-eye 3D imaging, and more specifically, to a naked-eye 3D imaging evaluation method and system based on feedback analysis. Background Technology

[0002] Glasses-free 3D technology typically uses parallax barriers, lenticular lenses, or other optical elements to separate the images for the left and right eyes and set different display images to achieve a glasses-free 3D effect. Glasses-free 3D display technology has seen some development and application in advertising screens, televisions, digital photo frames, game consoles, and other devices.

[0003] However, for some naked-eye 3D display platforms in exhibition halls, due to the presence of multiple users simultaneously performing 3D video analysis and display, and the large amount of naked-eye 3D content, the conversion is prone to excessive delays. Existing technologies are unable to achieve smooth synchronous display of multi-terminal image imaging, and the experience is poor for multi-user applications. In addition, existing technologies lack user interaction display and feedback analysis, making it difficult to dynamically adjust and display 3D content transmission and conversion. Summary of the Invention

[0004] This invention overcomes the shortcomings of the prior art and proposes a naked-eye 3D imaging evaluation method and system based on feedback analysis.

[0005] The first aspect of this invention provides a naked-eye 3D imaging evaluation method based on feedback analysis, comprising:

[0006] S101: Establishes a dedicated network connection between the cloud and the 3D display terminal, and stores the initial display data in the cloud;

[0007] S102: Users interact through the 3D display terminal. The cloud transforms the initial display data into naked-eye 3D content, which is then displayed through the 3D display terminal. Within one interaction cycle, the user's interaction information for each initial display data is collected, and an interaction matrix is ​​generated based on the interaction information.

[0008] S103: Based on initial display data, the camera module simulates the acquisition of left and right image data from the 3D display end by both eyes according to the user's viewing distance. The preset pixel matrix is ​​used as a moving window to move from the left and right image data. The color value histogram of the preset pixel matrix in the left and right image data is calculated based on each movement. The visual difference between the two eyes in the left and right image data is analyzed by the feedback of the color value histogram. The imaging weight is set based on the visual difference assessment.

[0009] S104: Perform feature decomposition through the interaction matrix and evaluate the differences in the interaction matrix. Based on the differences, classify the initial display data. Based on the classification results and combined with the imaging weights, set the priority group of the initial display data.

[0010] S105: By using priority groups, cache settings are configured for the initial display data based on the cloud to generate a 3D display caching scheme.

[0011] In this solution, S101 specifically refers to:

[0012] Establish a dedicated network connection between the cloud and multiple 3D display terminals, collect user interaction information in real time through the 3D display terminals and transmit it to the cloud for storage, and store the initial display data based on the cloud storage.

[0013] In this solution, S102 specifically refers to:

[0014] Define an interaction cycle and divide it into multiple time nodes;

[0015] Within an interaction cycle, the initial display data corresponding to the user interaction process is determined through the 3D display terminal, and the corresponding interaction information is collected;

[0016] Interaction information includes the interaction frequency, number of views, and viewing time at each time point in an initial display of data;

[0017] An interaction matrix is ​​constructed with time nodes as the first dimension and interactive information as the second dimension.

[0018] In this solution, S103 specifically refers to:

[0019] The system obtains the user's binocular distance and viewing distance, and sets up two camera devices through the camera module to collect image data from the 3D display end, thus obtaining left and right image data.

[0020] Set a 3×3 pixel matrix as the moving window, move the window over the left image data, calculate the corresponding color histogram based on the pixel matrix in each move, and extract color features through the color histogram. After the move is over, calculate the entire left image data.

[0021] Based on the color features extracted from each movement, a left image feature set is formed;

[0022] Color features are extracted from the right image to form a feature set for the right image;

[0023] The feature differences are calculated by selecting the extracted color features from the left image feature set and the right image feature set respectively, resulting in multiple difference values;

[0024] The average value of multiple difference values ​​is obtained by averaging, and the imaging weight is set based on the visual difference value.

[0025] In this solution, S104 specifically refers to:

[0026] Perform eigenvalue decomposition on all interaction matrices to obtain eigenvalues ​​and eigenvectors. Calculate the difference between interaction matrices based on the eigenvalues ​​and eigenvectors of each interaction matrix to obtain the matrix difference value between every two interaction matrices.

[0027] Based on the matrix difference value, all interaction matrices are grouped so that the maximum matrix difference value within the same group does not exceed the preset difference, resulting in multiple groups of interaction matrices.

[0028] Multiple sets of display data are obtained by mapping multiple sets of interaction matrices to the categories of the initial display data;

[0029] Calculate the average imaging weight corresponding to the initial display data in each group of display data, and set the priority of each group of display data based on the average weight to obtain the priority group.

[0030] In this solution, S105 specifically refers to:

[0031] By using priority groups, a 3D display caching scheme is generated based on the cloud-based caching settings for the initial display data.

[0032] In the 3D display caching solution, the initial display data of the highest priority is transformed in real time and stored in the cache list via the cloud;

[0033] Preload the initial display data based on the second priority and set up a real-time queue task.

[0034] A second aspect of the present invention also provides a naked-eye 3D imaging evaluation system based on feedback analysis. The system includes a memory and a processor. The memory includes a naked-eye 3D imaging evaluation program based on feedback analysis. When executed by the processor, the naked-eye 3D imaging evaluation program based on feedback analysis performs the following steps:

[0035] S101: Establishes a dedicated network connection between the cloud and the 3D display terminal, and stores the initial display data in the cloud;

[0036] S102: Users interact through the 3D display terminal. The cloud transforms the initial display data into naked-eye 3D content, which is then displayed through the 3D display terminal. Within one interaction cycle, the user's interaction information for each initial display data is collected, and an interaction matrix is ​​generated based on the interaction information.

[0037] S103: Based on initial display data, the camera module simulates the acquisition of left and right image data from the 3D display end by both eyes according to the user's viewing distance. The preset pixel matrix is ​​used as a moving window to move from the left and right image data. The color value histogram of the preset pixel matrix in the left and right image data is calculated based on each movement. The visual difference between the two eyes in the left and right image data is analyzed by the feedback of the color value histogram. The imaging weight is set based on the visual difference assessment.

[0038] S104: Perform feature decomposition through the interaction matrix and evaluate the differences in the interaction matrix. Based on the differences, classify the initial display data. Based on the classification results and combined with the imaging weights, set the priority group of the initial display data.

[0039] S105: By using priority groups, cache settings are configured for the initial display data based on the cloud to generate a 3D display caching scheme.

[0040] A third aspect of the present invention also provides a computer-readable storage medium comprising a naked-eye 3D imaging evaluation program based on feedback analysis, wherein when the naked-eye 3D imaging evaluation program based on feedback analysis is executed by a processor, it implements the steps of the naked-eye 3D imaging evaluation method based on feedback analysis as described in any of the preceding claims.

[0041] This invention discloses a naked-eye 3D imaging evaluation method and system based on feedback analysis, relating to the field of 3D display technology. A dedicated network is established between the cloud and the 3D terminal to store and convert initial display data into naked-eye 3D content in real time. Based on feedback analysis, on the one hand, an interaction matrix is ​​generated through user interaction behavior to analyze differences in operational characteristics; on the other hand, binocular vision simulation technology is used to analyze the color histogram distribution of the left and right views using a moving pixel window to quantify imaging weights. The system performs feature decomposition on the interaction matrix, establishes content priority groups based on imaging weights, and dynamically adjusts the cloud caching strategy to improve 3D display efficiency. This invention effectively improves user experience and optimizes imaging quality, effectively enhances the screen interaction efficiency of naked-eye 3D display devices, and improves the multi-user display efficiency of 3D display terminals. Attached Figure Description

[0042] Figure 1 A flowchart of a naked-eye 3D imaging evaluation method based on feedback analysis according to the present invention is shown;

[0043] Figure 2 A block diagram of a naked-eye 3D imaging evaluation system based on feedback analysis according to the present invention is shown. Detailed Implementation

[0044] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0045] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0046] Figure 1 A flowchart of a naked-eye 3D imaging evaluation method based on feedback analysis according to the present invention is shown.

[0047] like Figure 1 As shown, the first aspect of the present invention provides a naked-eye 3D imaging evaluation method based on feedback analysis, comprising:

[0048] S101: Establishes a dedicated network connection between the cloud and the 3D display terminal, and stores the initial display data in the cloud;

[0049] S102: Users interact through the 3D display terminal. The cloud transforms the initial display data into naked-eye 3D content, which is then displayed through the 3D display terminal. Within one interaction cycle, the user's interaction information for each initial display data is collected, and an interaction matrix is ​​generated based on the interaction information.

[0050] S103: Based on initial display data, the camera module simulates the acquisition of left and right image data from the 3D display end by both eyes according to the user's viewing distance. The preset pixel matrix is ​​used as a moving window to move from the left and right image data. The color value histogram of the preset pixel matrix in the left and right image data is calculated based on each movement. The visual difference between the two eyes in the left and right image data is analyzed by the feedback of the color value histogram. The imaging weight is set based on the visual difference assessment.

[0051] S104: Perform feature decomposition through the interaction matrix and evaluate the differences in the interaction matrix. Based on the differences, classify the initial display data. Based on the classification results and combined with the imaging weights, set the priority group of the initial display data.

[0052] S105: By using priority groups, cache settings are configured for the initial display data based on the cloud to generate a 3D display caching scheme.

[0053] It should be noted that the 3D display terminal of this invention includes a glasses-free 3D 8K display screen, which can convert ordinary 2D videos or traditional 3D (which requires wearing glasses) content into glasses-free 3D 8K content. The glasses-free 3D content can be directly played through devices such as HDMI interface, media player, and video camera. It can be widely used in commercial glasses-free 3D advertising screens, glasses-free 3D TVs, glasses-free 3D digital photo frames, glasses-free 3D game consoles, etc. For home glasses-free 3D displays, multimedia content on mobile phones can be converted through an app, stored in the cloud, and further displayed on the 3D display terminal.

[0054] According to an embodiment of the present invention, step S101 specifically includes:

[0055] Establish a dedicated network connection between the cloud and multiple 3D display terminals, collect user interaction information in real time through the 3D display terminals and transmit it to the cloud for storage, and store the initial display data based on the cloud storage.

[0056] It should be noted that the cloud is used for complex interactive data analysis, converting ordinary media content into glasses-free 3D content, and storing the initial display data. The initial display data refers to ordinary media content, such as 2D videos and images, which are then converted into glasses-free 3D content. For large files of display data, they can be divided into multiple parts for display, generating multiple initial display data sets for storage.

[0057] 3D display terminals are generally 3D exhibition halls, 3D advertising screens, glasses-free 3D TVs, glasses-free 3D digital photo frames, glasses-free 3D game consoles, etc. In exhibition halls and multimedia applications, there is a large amount of display data, which is generally ordinary 2D content. This invention represents the initial display data. For larger data display platforms, such as multi-user digital sand tables and multimedia exhibition hall applications, there is a lot of display data, and the 3D display content generally needs to be dynamically set. Therefore, it is necessary to store a large amount of initial display data in a database and dynamically convert and present it at any time.

[0058] According to an embodiment of the present invention, step S102 specifically includes:

[0059] Define an interaction cycle and divide it into multiple time nodes;

[0060] Within an interaction cycle, the initial display data corresponding to the user interaction process is determined through the 3D display terminal, and the corresponding interaction information is collected;

[0061] Interaction information includes the interaction frequency, number of views, and viewing time at each time point in an initial display of data;

[0062] An interaction matrix is ​​constructed with time nodes as the first dimension and interactive information as the second dimension.

[0063] It should be noted that the initial display data includes multiple data points, each corresponding to interactive information. Each interactive information corresponds to an interaction matrix. It's worth mentioning that for different types of 3D glasses-free display content, such as those displayed on exhibition hall displays and multimedia glasses-free displays, users typically need to browse, click, and view multiple times. The corresponding behavioral characteristic data is quite complex and involves different display data. Therefore, this invention subdivides the display data, stores it as initial display data, and analyzes user interaction information based on each initial display data point. This information is stored in the form of an interaction matrix. Subsequently, a rapid comparison process using the interaction matrix for each type of initial display data can establish the correlation between display data and user behavior, and uncover user interest display data. Later, through the interaction matrix, precise comparative analysis of corresponding interactive characteristics can be achieved.

[0064] According to an embodiment of the present invention, step S103 specifically includes:

[0065] The system obtains the user's binocular distance and viewing distance, and sets up two camera devices through the camera module to collect image data from the 3D display end, thus obtaining left and right image data.

[0066] Set a 3×3 pixel matrix as the moving window, move the window over the left image data, calculate the corresponding color histogram based on the pixel matrix in each move, and extract color features through the color histogram. After the move is over, calculate the entire left image data.

[0067] Based on the color features extracted from each movement, a left image feature set is formed;

[0068] Color features are extracted from the right image to form a feature set for the right image;

[0069] The feature differences are calculated by selecting the extracted color features from the left image feature set and the right image feature set respectively, resulting in multiple difference values;

[0070] The average value of multiple difference values ​​is obtained by averaging, and the imaging weight is set based on the visual difference value.

[0071] It should be noted that the binocular distance refers to the distance between the user's left and right eyes, the viewing distance is the distance from the user's eyes to the 3D display screen, and the left and right image data include left and right image data. The camera module includes two camera devices used to simulate human eye image acquisition and analysis, and the distance between the camera devices is set to the user's binocular distance. Simultaneously, the size of the acquired image is adjusted to match the actual image size obtained at the viewing distance. Feature difference calculation can be based on converting features into feature vectors and calculating based on Euclidean distance. The imaging weight is proportional to the visual difference value.

[0072] The camera module, based on the user's viewing distance, simulates binoculars to capture left and right image data from the 3D display. Specifically, it captures the user's view while viewing an initial display. Multiple views can be captured, including both left and right views. During window movement, movement starts from the top left corner of the image and proceeds line by line, with a unit of 3 pixels (the preset pixel matrix size). The feature set analysis process for the left and right images is consistent.

[0073] It's worth mentioning that naked-eye 3D technology utilizes parallax barriers and lenticular lenses to separate the images for the left and right eyes. This ensures the left eye sees only the corresponding left-view image, while the right eye sees the right-view image, thus simulating stereoscopic vision in the real world. Therefore, the difference between the left and right eye images can, to some extent, determine the imaging effect. In this invention, a pixel matrix is ​​used to compare and analyze the left and right images, calculating the differences through the color histogram within the pixel matrix. This method enables rapid analysis of pixel differences between the left and right eyes and real-time imaging evaluation. Traditional comparison methods often compare pixels one by one, resulting in significant redundant calculations and high computational costs. Furthermore, camera analysis is subject to pixel errors. Therefore, using a pixel matrix for extraction and analysis enables effective and rapid binocular difference analysis, further facilitating effective classification and caching of displayed data in subsequent steps.

[0074] Imaging weights are used to reflect imaging quality; the greater the visual difference, the better the quality.

[0075] According to an embodiment of the present invention, step S104 specifically includes:

[0076] Perform eigenvalue decomposition on all interaction matrices to obtain eigenvalues ​​and eigenvectors. Calculate the difference between interaction matrices based on the eigenvalues ​​and eigenvectors of each interaction matrix to obtain the matrix difference value between every two interaction matrices.

[0077] Based on the matrix difference value, all interaction matrices are grouped so that the maximum matrix difference value within the same group does not exceed the preset difference, resulting in multiple groups of interaction matrices.

[0078] Multiple sets of display data are obtained by mapping multiple sets of interaction matrices to the categories of the initial display data;

[0079] Calculate the average imaging weight corresponding to the initial display data in each group of display data, and set the priority of each group of display data based on the average weight to obtain the priority group.

[0080] It should be noted that matrix difference (similarity) is determined by combining the difference in eigenvalues ​​and the Euclidean distance between eigenvectors to assess the dissimilarity of two interaction matrices. Specifically, it is equal to the weighted average of the difference in eigenvalues ​​and the Euclidean distance between eigenvectors, yielding the matrix difference value. One initial display data set corresponds to one interaction matrix. Priority groups consist of multiple sets of display data that include priority information. Each interaction matrix corresponds to independent eigenvalues ​​and eigenvectors.

[0081] According to an embodiment of the present invention, step S105 specifically includes:

[0082] By using priority groups, a 3D display caching scheme is generated based on the cloud-based caching settings for the initial display data.

[0083] In the 3D display caching solution, the initial display data of the highest priority is transformed in real time and stored in the cache list via the cloud;

[0084] Preload the initial display data based on the second priority and set up a real-time queue task.

[0085] It should be noted that, based on priority groups, layered caching settings are implemented for display data groups with different priorities. Dynamic optimization of the caching mechanism and screen data transmission scheme can greatly improve the 3D glasses-free platform experience for multiple users. At the same time, it reduces data conversion pressure, effectively improves the smoothness of the screen on multiple 3D glasses-free display terminals, improves the real-time screen output capability of 3D display, and thus improves the overall user experience of 3D glasses-free exhibition halls (or multimedia 3D display terminals).

[0086] According to an embodiment of the present invention, it further includes:

[0087] The variance of the imaging weights corresponding to all initial display data is calculated to obtain the first variance;

[0088] The variance of the eigenvalues ​​corresponding to all interaction matrices is calculated to obtain the second variance;

[0089] Based on the first variance and the second variance, the consistency between user behavior characteristics and data imaging effects is determined;

[0090] If the consistency is lower than expected, then the feature vectors of all interaction matrices are used as clustering samples.

[0091] The K-means algorithm is used to group the clustered samples. K cluster centers are set, and the distance between feature vectors is measured by Euclidean distance. The clustering is repeated to form K groups of data and the center centers are updated repeatedly until the center centers no longer move, and the clustering ends.

[0092] Based on the clustering results, the initial display data is grouped and mapped to obtain multiple sets of display data. The average imaging weight of each set of display data is calculated to determine the priority, resulting in priority groups.

[0093] It should be noted that in 3D naked-eye video conversion and display applications based on multi-user, big data video libraries, it is difficult to conduct accurate group analysis of the 3D content behavior characteristics of multiple users. For example, in a multi-user urban 3D naked-eye display platform or exhibition hall, different users usually have different display data characteristics and user interaction behavior characteristics. When the user group is large, existing technologies are difficult to conduct accurate analysis.

[0094] It's worth noting that for some glasses-free 3D display platforms in exhibition halls, due to the simultaneous 3D video analysis and display by multiple users, the large amount of glasses-free 3D content, and the tendency for excessive latency during conversion, existing technologies struggle to achieve smooth, synchronized display of images from multiple terminals. Furthermore, the user experience is poor in multi-user applications. Moreover, existing technologies lack user interaction feedback analysis, making it difficult to dynamically adjust and display 3D content transmission and conversion. Particularly in scenarios such as data sandboxes and multimedia exhibition halls, existing technologies have low display efficiency for glasses-free 3D displays with real-time video conversion, resulting in a poor multi-user experience and difficulty in user interaction matching, thus hindering the improvement of glasses-free 3D viewing efficiency in various scenarios.

[0095] Based on the above problems, on the one hand, this invention uses the matrix difference value of the interaction matrix for group setting, which is suitable for analysis when the number of user groups is small. On the other hand, when the user group is large, the consistency between the corresponding user interaction and display characteristics is low, and precise grouping analysis is required. Therefore, this invention analyzes the imaging weight of the display data and the discreteness of the interaction characteristics, judges the consistency between the display data and the interaction characteristics of the user group by calculating the variance, and judges whether clustering of the interaction data and fine grouping of the user group are required based on the discreteness (consistency). In this way, a priority group with higher accuracy is selected and a caching scheme suitable for multiple user groups is set.

[0096] In the example, if the consistency is lower than expected, that is, both variance values ​​are lower than expected.

[0097] According to an embodiment of the present invention, it further includes:

[0098] In the second interaction cycle, the cloud and 3D display terminal are interacted based on the 3D display caching solution;

[0099] Based on the second interaction cycle, analyze the user's interaction matrix, calculate the corresponding matrix difference value, and calculate the mean of the matrix difference value to obtain the second mean.

[0100] Calculate the mean of the matrix differences during the first interaction cycle to obtain the first mean;

[0101] If the second mean is greater than the first mean and the absolute difference is greater than the threshold, then calculate the display transmission efficiency of the first interaction cycle and the second interaction cycle.

[0102] The efficiency of the 3D display caching scheme is evaluated by comparing the absolute difference with the transmission efficiency in two interaction cycles.

[0103] It should be noted that the matrix difference value is an indicator reflecting the differences in user interaction characteristics. When the differences in interaction characteristics among user groups are significant, the complexity of the displayed data content will increase accordingly. Therefore, when the interaction differences are large between two periods, the efficiency of the 3D display caching scheme can be efficiently evaluated. Based on this, this invention identifies changes in interaction differences by periodically comparing the matrix difference value, and further analyzes the caching efficiency by comparing the display transmission efficiency.

[0104] Display transmission efficiency is specifically an efficiency information that comprehensively analyzes information such as the conversion time of the user's initial display data, screen latency, and data transmission time between the cloud and the 3D display terminal. Generally speaking, the larger the absolute difference and the higher the display transmission efficiency, the more efficient the caching solution is and the more suitable it is for the current 3D display platform.

[0105] The absolute difference is the absolute difference between the first mean and the second mean.

[0106] Figure 2 A block diagram of a naked-eye 3D imaging evaluation system based on feedback analysis according to the present invention is shown.

[0107] A second aspect of the present invention also provides a naked-eye 3D imaging evaluation system 2 based on feedback analysis. The system includes a memory 21 and a processor 22. The memory 21 includes a naked-eye 3D imaging evaluation program based on feedback analysis. When the processor 22 executes the naked-eye 3D imaging evaluation program based on feedback analysis, it performs the following steps:

[0108] S101: Establishes a dedicated network connection between the cloud and the 3D display terminal, and stores the initial display data in the cloud;

[0109] S102: Users interact through the 3D display terminal. The cloud transforms the initial display data into naked-eye 3D content, which is then displayed through the 3D display terminal. Within one interaction cycle, the user's interaction information for each initial display data is collected, and an interaction matrix is ​​generated based on the interaction information.

[0110] S103: Based on initial display data, the camera module simulates the acquisition of left and right image data from the 3D display end by both eyes according to the user's viewing distance. The preset pixel matrix is ​​used as a moving window to move from the left and right image data. The color value histogram of the preset pixel matrix in the left and right image data is calculated based on each movement. The visual difference between the two eyes in the left and right image data is analyzed by the feedback of the color value histogram. The imaging weight is set based on the visual difference assessment.

[0111] S104: Perform feature decomposition through the interaction matrix and evaluate the differences in the interaction matrix. Based on the differences, classify the initial display data. Based on the classification results and combined with the imaging weights, set the priority group of the initial display data.

[0112] S105: By using priority groups, cache settings are configured for the initial display data based on the cloud to generate a 3D display caching scheme.

[0113] It should be noted that the 3D display terminal of this invention includes a glasses-free 3D 8K display screen, which can convert ordinary 2D videos or traditional 3D (which requires wearing glasses) content into glasses-free 3D 8K content. The glasses-free 3D content can be directly played through devices such as HDMI interface, media player, and video camera. It can be widely used in commercial glasses-free 3D advertising screens, glasses-free 3D TVs, glasses-free 3D digital photo frames, glasses-free 3D game consoles, etc. For home glasses-free 3D displays, multimedia content on mobile phones can be converted through an app, stored in the cloud, and further displayed on the 3D display terminal.

[0114] According to an embodiment of the present invention, step S101 specifically includes:

[0115] Establish a dedicated network connection between the cloud and multiple 3D display terminals, collect user interaction information in real time through the 3D display terminals and transmit it to the cloud for storage, and store the initial display data based on the cloud storage.

[0116] It should be noted that the cloud is used for complex interactive data analysis, converting ordinary media content into glasses-free 3D content, and storing the initial display data. The initial display data refers to ordinary media content, such as 2D videos and images, which are then converted into glasses-free 3D content. For large files of display data, they can be divided into multiple parts for display, generating multiple initial display data sets for storage.

[0117] 3D display terminals are generally 3D exhibition halls, 3D advertising screens, glasses-free 3D TVs, glasses-free 3D digital photo frames, glasses-free 3D game consoles, etc. In exhibition halls and multimedia applications, there is a large amount of display data, which is generally ordinary 2D content. This invention represents the initial display data. For larger data display platforms, such as multi-user digital sand tables and multimedia exhibition hall applications, there is a lot of display data, and the 3D display content generally needs to be dynamically set. Therefore, it is necessary to store a large amount of initial display data in a database and dynamically convert and present it at any time.

[0118] According to an embodiment of the present invention, step S102 specifically includes:

[0119] Define an interaction cycle and divide it into multiple time nodes;

[0120] Within an interaction cycle, the initial display data corresponding to the user interaction process is determined through the 3D display terminal, and the corresponding interaction information is collected;

[0121] Interaction information includes the interaction frequency, number of views, and viewing time at each time point in an initial display of data;

[0122] An interaction matrix is ​​constructed with time nodes as the first dimension and interactive information as the second dimension.

[0123] It should be noted that the initial display data includes multiple data points, each corresponding to interactive information. Each interactive information corresponds to an interaction matrix. It's worth mentioning that for different types of 3D glasses-free display content, such as those displayed on exhibition hall displays and multimedia glasses-free displays, users typically need to browse, click, and view multiple times. The corresponding behavioral characteristic data is quite complex and involves different display data. Therefore, this invention subdivides the display data, stores it as initial display data, and analyzes user interaction information based on each initial display data point. This information is stored in the form of an interaction matrix. Subsequently, a rapid comparison process using the interaction matrix for each type of initial display data can establish the correlation between display data and user behavior, and uncover user interest display data. Later, through the interaction matrix, precise comparative analysis of corresponding interactive characteristics can be achieved.

[0124] According to an embodiment of the present invention, step S103 specifically includes:

[0125] The system obtains the user's binocular distance and viewing distance, and sets up two camera devices through the camera module to collect image data from the 3D display end, thus obtaining left and right image data.

[0126] Set a 3×3 pixel matrix as the moving window, move the window over the left image data, calculate the corresponding color histogram based on the pixel matrix in each move, and extract color features through the color histogram. After the move is over, calculate the entire left image data.

[0127] Based on the color features extracted from each movement, a left image feature set is formed;

[0128] Color features are extracted from the right image to form a feature set for the right image;

[0129] The feature differences are calculated by selecting the extracted color features from the left image feature set and the right image feature set respectively, resulting in multiple difference values;

[0130] The average value of multiple difference values ​​is obtained by averaging, and the imaging weight is set based on the visual difference value.

[0131] It should be noted that the binocular distance refers to the distance between the user's left and right eyes, the viewing distance is the distance from the user's eyes to the 3D display screen, and the left and right image data include left and right image data. The camera module includes two camera devices used to simulate human eye image acquisition and analysis, and the distance between the camera devices is set to the user's binocular distance. Simultaneously, the size of the acquired image is adjusted to match the actual image size obtained at the viewing distance. Feature difference calculation can be based on converting features into feature vectors and calculating based on Euclidean distance. The imaging weight is proportional to the visual difference value.

[0132] The camera module, based on the user's viewing distance, simulates binoculars to capture left and right image data from the 3D display. Specifically, it captures the user's view while viewing an initial display. Multiple views can be captured, including both left and right views. During window movement, movement starts from the top left corner of the image and proceeds line by line, with a unit of 3 pixels (the preset pixel matrix size). The feature set analysis process for the left and right images is consistent.

[0133] It's worth mentioning that naked-eye 3D technology utilizes parallax barriers and lenticular lenses to separate the images for the left and right eyes. This ensures the left eye sees only the corresponding left-view image, while the right eye sees the right-view image, thus simulating stereoscopic vision in the real world. Therefore, the difference between the left and right eye images can, to some extent, determine the imaging effect. In this invention, a pixel matrix is ​​used to compare and analyze the left and right images, calculating the differences through the color histogram within the pixel matrix. This method enables rapid analysis of pixel differences between the left and right eyes and real-time imaging evaluation. Traditional comparison methods often compare pixels one by one, resulting in significant redundant calculations and high computational costs. Furthermore, camera analysis is subject to pixel errors. Therefore, using a pixel matrix for extraction and analysis enables effective and rapid binocular difference analysis, further facilitating effective classification and caching of displayed data in subsequent steps.

[0134] Imaging weights are used to reflect imaging quality; the greater the visual difference, the better the quality.

[0135] According to an embodiment of the present invention, step S104 specifically includes:

[0136] Perform eigenvalue decomposition on all interaction matrices to obtain eigenvalues ​​and eigenvectors. Calculate the difference between interaction matrices based on the eigenvalues ​​and eigenvectors of each interaction matrix to obtain the matrix difference value between every two interaction matrices.

[0137] Based on the matrix difference value, all interaction matrices are grouped so that the maximum matrix difference value within the same group does not exceed the preset difference, resulting in multiple groups of interaction matrices.

[0138] Multiple sets of display data are obtained by mapping multiple sets of interaction matrices to the categories of the initial display data;

[0139] Calculate the average imaging weight corresponding to the initial display data in each group of display data, and set the priority of each group of display data based on the average weight to obtain the priority group.

[0140] It should be noted that matrix difference (similarity) is determined by combining the difference in eigenvalues ​​and the Euclidean distance between eigenvectors to assess the dissimilarity of two interaction matrices. Specifically, it is equal to the weighted average of the difference in eigenvalues ​​and the Euclidean distance between eigenvectors, yielding the matrix difference value. One initial display data set corresponds to one interaction matrix. Priority groups consist of multiple sets of display data that include priority information. Each interaction matrix corresponds to independent eigenvalues ​​and eigenvectors.

[0141] According to an embodiment of the present invention, step S105 specifically includes:

[0142] By using priority groups, a 3D display caching scheme is generated based on the cloud-based caching settings for the initial display data.

[0143] In the 3D display caching solution, the initial display data of the highest priority is transformed in real time and stored in the cache list via the cloud;

[0144] Preload the initial display data based on the second priority and set up a real-time queue task.

[0145] It should be noted that, based on priority groups, layered caching settings are implemented for display data groups with different priorities. Dynamic optimization of the caching mechanism and screen data transmission scheme can greatly improve the 3D glasses-free platform experience for multiple users. At the same time, it reduces data conversion pressure, effectively improves the smoothness of the screen on multiple 3D glasses-free display terminals, improves the real-time screen output capability of 3D display, and thus improves the overall user experience of 3D glasses-free exhibition halls (or multimedia 3D display terminals).

[0146] A third aspect of the present invention also provides a computer-readable storage medium comprising a naked-eye 3D imaging evaluation program based on feedback analysis, wherein when the naked-eye 3D imaging evaluation program based on feedback analysis is executed by a processor, it implements the steps of the naked-eye 3D imaging evaluation method based on feedback analysis as described in any of the preceding claims.

[0147] This invention discloses a naked-eye 3D imaging evaluation method and system based on feedback analysis, relating to the field of 3D display technology. A dedicated network is established between the cloud and the 3D terminal to store and convert initial display data into naked-eye 3D content in real time. Based on feedback analysis, on the one hand, an interaction matrix is ​​generated through user interaction behavior to analyze differences in operational characteristics; on the other hand, binocular vision simulation technology is used to analyze the color histogram distribution of the left and right views using a moving pixel window to quantify imaging weights. The system performs feature decomposition on the interaction matrix, establishes content priority groups based on imaging weights, and dynamically adjusts the cloud caching strategy to improve 3D display efficiency. This invention effectively improves user experience and optimizes imaging quality, effectively enhances the screen interaction efficiency of naked-eye 3D display devices, and improves the multi-user display efficiency of 3D display terminals.

[0148] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0149] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0150] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0151] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0153] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A naked-eye 3D imaging evaluation method based on feedback analysis, characterized in that, include: S101: Establishes a dedicated network connection between the cloud and the 3D display terminal, and stores the initial display data in the cloud; S102: Users interact through the 3D display terminal. The cloud transforms the initial display data into naked-eye 3D content, which is then displayed through the 3D display terminal. Within one interaction cycle, the user's interaction information for each initial display data is collected, and an interaction matrix is ​​generated based on the interaction information. S103: Based on initial display data, the camera module simulates the acquisition of left and right image data from the 3D display end by both eyes according to the user's viewing distance. The preset pixel matrix is ​​used as a moving window to move from the left and right image data. The color value histogram of the preset pixel matrix in the left and right image data is calculated based on each movement. The visual difference between the two eyes in the left and right image data is analyzed by the feedback of the color value histogram. The imaging weight is set based on the visual difference assessment. S104: Perform feature decomposition through the interaction matrix and evaluate the differences in the interaction matrix. Based on the differences, classify the initial display data. Based on the classification results and combined with the imaging weights, set the priority group of the initial display data. S105: By using priority groups, cache settings are configured for the initial display data based on the cloud to generate a 3D display caching scheme; Specifically, S104 is as follows: Perform eigenvalue decomposition on all interaction matrices to obtain eigenvalues ​​and eigenvectors. Calculate the difference between interaction matrices based on the eigenvalues ​​and eigenvectors of each interaction matrix to obtain the matrix difference value between every two interaction matrices. Based on the matrix difference value, all interaction matrices are grouped so that the maximum matrix difference value within the same group does not exceed the preset difference, resulting in multiple groups of interaction matrices. Multiple sets of display data are obtained by mapping multiple sets of interaction matrices to the categories of the initial display data; Calculate the average imaging weight corresponding to the initial display data in each group of display data, and set the priority of each group of display data based on the average weight to obtain the priority group.

2. The naked-eye 3D imaging evaluation method based on feedback analysis according to claim 1, characterized in that, Specifically, S101 is as follows: Establish a dedicated network connection between the cloud and multiple 3D display terminals, collect user interaction information in real time through the 3D display terminals and transmit it to the cloud for storage, and store the initial display data based on the cloud storage.

3. The naked-eye 3D imaging evaluation method based on feedback analysis according to claim 1, characterized in that, Specifically, S102 is as follows: Define an interaction cycle and divide it into multiple time nodes; Within an interaction cycle, the initial display data corresponding to the user interaction process is determined through the 3D display terminal, and the corresponding interaction information is collected; Interaction information includes the interaction frequency, number of views, and viewing time at each time point in an initial display of data; An interaction matrix is ​​constructed with time nodes as the first dimension and interactive information as the second dimension.

4. The naked-eye 3D imaging evaluation method based on feedback analysis according to claim 1, characterized in that, Specifically, S103 is as follows: The system obtains the user's binocular distance and viewing distance, and sets up two camera devices through the camera module to collect image data from the 3D display end, thus obtaining left and right image data. Set a 3×3 pixel matrix as the moving window, move the window over the left image data, calculate the corresponding color histogram based on the pixel matrix in each move, and extract color features through the color histogram. After the move is over, calculate the entire left image data. Based on the color features extracted from each movement, a left image feature set is formed; Color features are extracted from the right image to form a feature set for the right image; The feature differences are calculated by selecting the extracted color features from the left image feature set and the right image feature set respectively, resulting in multiple difference values; The average value of multiple difference values ​​is obtained by averaging, and the imaging weight is set based on the visual difference value.

5. The naked-eye 3D imaging evaluation method based on feedback analysis according to claim 1, characterized in that, Specifically, S105 is as follows: By using priority groups, a 3D display caching scheme is generated based on the cloud-based caching settings for the initial display data. In the 3D display caching solution, the initial display data of the highest priority is transformed in real time and stored in the cache list via the cloud; Preload the initial display data based on the second priority and set up a real-time queue task.

6. A naked-eye 3D imaging evaluation system based on feedback analysis, characterized in that, The system includes: a memory and a processor. The memory includes a naked-eye 3D imaging evaluation program based on feedback analysis. When the naked-eye 3D imaging evaluation program based on feedback analysis is executed by the processor, it implements the steps of the naked-eye 3D imaging evaluation method based on feedback analysis as described in claim 1.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a naked-eye 3D imaging evaluation program based on feedback analysis, which, when executed by a processor, implements the steps of the naked-eye 3D imaging evaluation method based on feedback analysis as described in any one of claims 1 to 5.

Images