Data disassembling and visual processing method and system for intelligent split shot

Abstract

The application discloses a smart split-screen data disassembling and visualization processing method and system, and belongs to the technical field of visualization processing, and comprises the following steps: script text analysis and extraction efficiency regulation, split-screen unit disassembly and two-dimensional regulation, split-screen structured processing and visual reference map generation, and visual reference map verification, sorting and complete conversion. The application obtains a script text to be processed, analyzes and regulates the main information extraction efficiency in combination with a standardized film language database, matches the information to be disassembled into split-screen units, regulates the disassembly accuracy and efficiency, structures the split screen, generates a visual reference map in combination with the role and scene data, verifies and sorts the split-screen visualization sequence, and completes the automatic conversion of the script text into split-screen visual data, thereby improving the data disassembly and visualization processing accuracy and solving the problem of low data disassembly and visualization processing accuracy in the prior art.

Description

Technical Field

[0001] This invention relates to the field of visualization processing technology, and in particular to a method and system for data decomposition and visualization processing of intelligent storyboards. Background Technology

[0002] Existing technologies for converting script text into storyboards generally rely heavily on human experience for storyboard design and production. This not only results in a cumbersome and time-consuming process, leading to low storyboard production efficiency and failing to meet the demands of rapid industrial production in film, animation, and short videos, but also places the quality of storyboards entirely on the creator's professional level. Inconsistent and non-standard use of camera language, coupled with a lack of professional cinematic logic in the selection of camera angles, framing, and camera movement, easily leads to problems such as chaotic narrative pacing and imprecise camera expression. Furthermore, existing technologies lack a connection between script text information and storyboard visual elements. The standardized and reusable mapping rules between elements cannot structurally extract and quantitatively decompose key information such as scenes, actions, dialogues, and emotions in the script. It is difficult to form standardized and unified storyboard unit data, and it is even more difficult to effectively transform text data into structured storyboard parameters such as shot type, duration, transition, and composition that can be recognized and executed by machines. Consequently, it is impossible to achieve the automated generation and visualization of visual content from text to storyboard. The overall level of intelligence and automation is extremely low, which seriously restricts the large-scale, standardized, and efficient development of storyboard production and results in low accuracy of data decomposition and visualization processing. Summary of the Invention

[0003] To address the low accuracy of data decomposition and visualization in existing technologies, this invention provides a method and system for intelligent storyboard data decomposition and visualization. The technical solution is as follows: On the one hand, a data decomposition and visualization method for intelligent storyboarding is provided. This method includes: Step 101, acquiring the script text to be processed, parsing the script text segment by segment based on a pre-built standardized film language database, and determining whether to perform main information extraction efficiency adjustment based on the results of segment-by-segment parsing to improve the efficiency of automatically extracting the main information of the script text to be processed; Step 102, matching the extracted main information of the script text to be processed with the standardized film language database, decomposing the script text to be processed into several independent storyboard units based on the matching results, determining whether to perform storyboard unit decomposition precision adjustment based on the time consumption of the storyboard unit decomposition within the storyboard unit, and if so, determining whether to perform decomposition after adjustment. If efficiency coordination is not performed, then it is directly determined whether to perform efficiency coordination control. Step 103: Each storyboard unit obtained from the decomposition is structured and converted into structured storyboard data that can be recognized by machines. The structured storyboard data is mapped and adapted with the character feature data and scene feature data corresponding to the script text to be processed to generate a visual reference map corresponding to each storyboard unit. Step 104: The visual reference map corresponding to each generated storyboard unit is standardized and verified. All visual reference maps are sorted to form a complete storyboard visualization sequence, completing the automated conversion from script text to storyboard visual data, and realizing the data decomposition and visualization processing of intelligent storyboards.

[0004] On the other hand, an intelligent storyboard data decomposition and visualization processing system is provided. This system includes: a script text parsing and extraction efficiency control module, a storyboard unit decomposition and dual-dimensional control module, a storyboard structured processing and visual reference image generation module, and a visual reference image verification, sorting, and complete conversion module. The script text parsing and extraction efficiency control module acquires the script text to be processed, parses it segment by segment based on a pre-built standardized film language database, and determines whether to perform main information extraction efficiency control based on the results of the segment-by-segment parsing to improve the efficiency of automatically extracting the main information of the script text. The storyboard unit decomposition and dual-dimensional control module matches the extracted main information of the script text to be processed with the standardized film language database, and decomposes the script text into several independent storyboard units based on the matching results. The system also controls the efficiency of decomposing the storyboard units within each storyboard unit. The system determines whether to perform segmentation accuracy adjustment for each segment. If so, it determines whether to perform segmentation efficiency coordination adjustment after adjustment; otherwise, it directly determines whether to perform segmentation efficiency coordination adjustment. The segmentation structuring and visual reference map generation module is used to perform structuring processing on each segmented segment, converting each segment into machine-recognizable structured segmentation data. It then maps and adapts the structured segmentation data with the character feature data and scene feature data corresponding to the script text to generate a visual reference map for each segment. The visual reference map verification, sorting, and complete conversion module is used to standardize and verify the visual reference map corresponding to each generated segment, sort all visual reference maps to form a complete segment visualization sequence, and complete the automated conversion from script text to segmentation visual data, realizing intelligent segmentation data decomposition and visualization processing.

[0005] Beneficial effects The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. The intelligent storyboard data decomposition and visualization processing method provided by this invention effectively solves the technical problems of low efficiency and poor adaptability in traditional script text analysis. It achieves standardized segment-by-segment analysis of the script text through a pre-constructed standardized film language database. Combined with dynamic adaptation and adjustment of the frequency of repeated information extraction and the granularity of segment analysis, it accurately matches the analysis needs of script texts of varying complexity, significantly improving the efficiency and accuracy of key information extraction. Specifically, by pre-setting three levels of analysis granularity (coarse, medium, and fine) and corresponding frequency reference ranges, when the frequency of repeated information extraction deviates from the reference range, the analysis granularity is automatically upgraded or downgraded. This avoids the omission of core information caused by coarse-grained analysis and the redundant calculations caused by fine-grained analysis. The analysis process can flexibly adapt to the distribution characteristics of core information such as scenes, actions, and dialogues in the script text, significantly reducing the time spent on ineffective analysis. This ensures that the extracted key information accurately supports subsequent storyboard decomposition work, laying an efficient and reliable foundation for the entire storyboard processing workflow. Simultaneously, it reduces the cost of manual intervention in analysis and improves the automation level of the analysis process.

[0006] 2. This invention effectively solves the technical pain points of traditional storyboard disassembly, such as difficulty in balancing accuracy and efficiency, excessive redundant storyboards, and unbalanced thread load, by achieving bidirectional collaborative optimization of storyboard unit disassembly accuracy and disassembly efficiency. Through the linkage control of storyboard disassembly time and redundant disassembly rate, and the dynamic adaptation of disassembly thread and cache reuse ratio, it ensures that storyboard disassembly meets standardization requirements while taking into account processing efficiency. On the one hand, by setting preset thresholds for the time interval and redundancy rate of storyboard decomposition, the system automatically adjusts the redundancy rate based on the actual decomposition time of a single storyboard unit. When the decomposition efficiency is too high, the redundancy rate is appropriately increased to ensure decomposition accuracy; when the decomposition efficiency is too low, the redundancy rate is appropriately decreased to increase speed, thus avoiding storyboard units without narrative significance from occupying computing resources. On the other hand, by monitoring the total number of decomposition threads in real time and combining the thread load, the system dynamically adjusts the data cache reuse ratio of decomposition storyboard units. Under low load, the cache reuse ratio is lowered to improve decomposition accuracy; under high load, the cache reuse ratio is increased to reduce thread computing pressure, achieving a reasonable allocation of decomposition thread resources. At the same time, through the logic of "prioritizing accuracy control and coordinating efficiency," the system ensures that the storyboard decomposition process can flexibly adapt to the actual computing situation. The decomposition of independent storyboard units not only conforms to standardized film language specifications but also effectively reduces invalid decomposition and redundant computing, significantly improving the overall quality and efficiency of storyboard decomposition.

[0007] 3. By significantly improving the accuracy and automation level of storyboard visualization, it effectively solves the technical problems of unreasonable character positions, screen cropping, non-standard visual reference images, and low degree of automation in traditional storyboard visualization. Through character position parameter mapping adaptation and adjustment, and standardized verification and sorting of visual reference images, it achieves full-process automated conversion from script text to storyboard visual data, forming a complete and standardized storyboard visualization sequence. Specifically, by collecting character interaction commands in real time and extracting interaction response times, the system dynamically adjusts the safety distance between the character and the edge of the screen based on preset response time thresholds. For fast interaction responses, the safety distance is increased to avoid screen cropping; for slow interaction responses, the safety distance is reduced to optimize screen space utilization, ensuring that the character's position accurately matches the storyboard intent, character interaction state, and scene logic. Simultaneously, through the coordinated operation of four functional modules, the storyboard units are structured into machine-recognizable data. Combining character and scene characteristics, character position adjustments are made, and after generating a visual reference image, standardized checks are performed to identify issues such as screen accuracy, character position, and scene adaptability. Finally, the images are sorted according to narrative logic to form a complete storyboard visualization sequence. This achieves a fully automated closed loop from script text analysis, storyboard decomposition, structured processing to visual presentation, eliminating the need for manual intervention in storyboard design and visual adjustments. This significantly reduces the labor costs of intelligent storyboard production, improves the standardization level and production efficiency of storyboard visualization, and provides an efficient, accurate, and standardized automated solution for storyboard production in film, animation, and other fields. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 Flowchart of the data decomposition and visualization method for intelligent storyboard provided in this application embodiment; Figure 2 A flowchart illustrating the role position parameter mapping and adaptation control process of the intelligent storyboard data decomposition and visualization method provided in this application embodiment; Figure 3 A schematic diagram of the data disassembly and visualization processing system for intelligent storyboards provided in this application embodiment. Detailed Implementation

[0010] Embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of the present disclosure are shown in the drawings, it should be understood that embodiments of the present disclosure may be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure.

[0011] It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure. In the description of the embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "this embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects.

[0012] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0013] like Figure 1 The diagram shown is a flowchart of the data decomposition and visualization processing method for intelligent storyboards provided in this application embodiment. The method includes the following steps: Step 101: Obtain the script text to be processed. Based on the pre-built standardized film language database, parse the script text to be processed segment by segment. Based on the results of segment-by-segment parsing, determine whether to perform main information extraction efficiency adjustment to improve the efficiency of automatically extracting the main information of the script text to be processed.

[0014] It should be understood that the process begins with acquiring the target script text to be processed, clarifying the core data source for subsequent automated processing, and laying the foundation for standardized analysis and information extraction. Next, a pre-built standardized film language database is invoked. This database has already standardized and categorized film-related language elements such as film terminology, scene descriptions, character behavior, dialogue logic, and camera language, establishing a unified and professional analytical reference standard to avoid analytical biases and recognition failures caused by inconsistent standards. Then, based on this standardized film language database, the script text is analyzed paragraph by paragraph. By matching each paragraph with standardized film language entries in the database, the language attributes, information types, structural features, and information density of each paragraph are accurately identified, and the analysis proceeds segment by segment. This approach refines the granularity of parsing and reduces the overall complexity of parsing, thereby improving the accuracy and reliability of the parsing results. Then, based on the parsing results obtained segment by segment, such as paragraph information density, structural complexity, and language standardization, a comprehensive judgment is made as to whether to perform main information extraction efficiency control operations. The parsing results drive the control decisions in reverse, ensuring that efficiency control is only triggered in scenarios with insufficient information extraction efficiency, high information redundancy, or chaotic text structure, thus avoiding waste of system resources and processing delays caused by ineffective control. Finally, when it is determined that control needs to be performed, an efficiency control strategy adapted to the corresponding paragraph characteristics is activated to optimize the information extraction logic, streamline redundant processing steps, and focus on high-value information screening. This achieves an efficiency improvement in automatically extracting the main information of the script text to be processed, while ensuring the accuracy and completeness of information extraction.

[0015] It should be further explained that the specific steps for determining whether to perform main information extraction efficiency adjustment based on the results of segment-by-segment parsing are as follows: The system presets the reference range for the frequency of repeated information extraction, and presets the paragraph parsing granularity level and corresponding granularity range. The paragraph parsing granularity level includes three levels: coarse-grained, medium-grained, and fine-grained. Coarse-grained corresponds to parsing the smallest text unit as a complete paragraph, medium-grained corresponds to parsing the smallest text unit as a sentence, and fine-grained corresponds to parsing the smallest text unit as a clause. The frequency of repeated information extraction is obtained. The frequency of repeated information extraction means the number of times the same core information is repeatedly extracted in different parsed text units. The core information is the scene information, action information, dialogue information, and emotion information in the script text to be processed. The parsed text unit is the smallest text unit corresponding to the parsing granularity of the current paragraph. The frequency of repeated information extraction is compared with a preset reference range for repeated information extraction, and the granularity of paragraph parsing is dynamically adjusted based on the comparison results. Specifically: If the frequency of repeated information extraction is within the preset reference range for repeated information extraction frequency, it is determined that the parsing granularity of the current paragraph is suitable for the parsing requirements of the current script text. The unparsed script text is then parsed paragraph by paragraph at the current parsing granularity. The preset reference range for repeated information extraction frequency represents the closed interval formed by the preset lower limit and the preset upper limit of the repeated information extraction frequency reference.

[0016] Dynamically adjusting the granularity of paragraph parsing based on the comparison results also includes: If the frequency of repeated information extraction is less than the preset lower limit of repeated information extraction frequency, the current paragraph is judged to be too coarse in terms of parsing granularity. The current frequency of repeated information extraction is input into the constructed extraction frequency-parsing granularity database, and a downgrade adjustment is output. The downgrade adjustment means that if the current granularity is coarse, it will be adjusted to medium granularity, and if the current granularity is medium granularity, it will be adjusted to fine granularity. After the adjustment, the unparsed script text will continue to be parsed paragraph by paragraph with the new parsing granularity. If the frequency of repeated information extraction exceeds the preset upper limit of repeated information extraction frequency, the current paragraph parsing granularity is determined to be too fine. The current repeated information extraction frequency is input into the constructed extraction frequency-parsing granularity database, and an upgrade adjustment is output. The upgrade adjustment means that if the current granularity is fine, it will be adjusted to medium granularity, and if the current granularity is medium, it will be adjusted to coarse granularity. After the adjustment, the unparsed script text will continue to be parsed paragraph by paragraph with the new parsing granularity.

[0017] In this embodiment, preliminary rules and parameters are first preset, clearly defining the reference range for the frequency of repeated information extraction. Simultaneously, the paragraph parsing granularity level and the specific granularity range corresponding to each level are preset. The paragraph parsing granularity level is clearly divided into three levels: coarse-grained, medium-grained, and fine-grained. The smallest text unit for parsing at each level is precisely defined: coarse-grained corresponds to a complete paragraph, medium-grained to a sentence, and fine-grained to a clause. This preset step provides a standardized reference and clear operational guidelines for subsequent repeated information frequency comparison and dynamic adjustment of parsing granularity. It effectively avoids the problems of lack of rules and chaotic granularity adjustment in subsequent operations, ensuring the standardization and consistency of granularity adjustment from the source, and laying the foundation for improving the overall parsing and information extraction efficiency. Subsequently, the frequency of repeated information extraction in the current parsing scenario is obtained. This frequency specifically refers to the number of times the same core information is repeatedly extracted in different parsed text units. Core information is clearly defined as four key types of information in the script text to be processed: scene information, action information, dialogue information, and emotional information. The parsed text unit corresponds to the smallest text unit determined by the current paragraph parsing granularity. By accurately obtaining the frequency of repeated information extraction, the redundancy of information extraction at the current parsing granularity can be intuitively reflected, providing core data support for subsequent granularity adjustments and avoiding the blindness of adjusting granularity based solely on experience. Focusing on the four types of core information ensures the relevance of frequency statistics, reduces interference from irrelevant information, and improves the effectiveness of frequency data. Next, the obtained frequency of repeated information extraction is precisely compared with a preset reference interval for repeated information extraction. This preset reference interval is a closed interval formed by the preset lower limit and the preset upper limit of the repeated information extraction frequency reference. Based on the comparison results, the paragraph parsing granularity is dynamically adjusted to achieve precise adaptation between the parsing granularity and the text parsing requirements, thereby optimizing information extraction efficiency and controlling information redundancy.Specifically, if the frequency of repeated information extraction falls within a preset reference range for repeated information extraction frequency, it is determined that the current paragraph parsing granularity is fully adapted to the parsing requirements of the current script text. In this case, the current parsing granularity remains unchanged, and the unparsed script text continues to be parsed paragraph by paragraph at this granularity. This judgment logic ensures that no invalid adjustments are made in scenarios where the parsing granularity is adapted, avoiding unnecessary waste of system resources due to unnecessary granularity switching, while maintaining the stability of the parsing rhythm and ensuring that the efficiency and accuracy of information extraction are not affected. If the frequency of repeated information extraction is less than the preset repeated information... If the extraction frequency is set to the lower limit, the current paragraph's parsing granularity is determined to be too coarse. In this case, the current extraction frequency of repeated information is input into a pre-built extraction frequency-parsing granularity database. The database then outputs a downgrade adjustment instruction. Specifically, if the current parsing granularity is coarse, it is adjusted to medium granularity; if it is medium granularity, it is adjusted to fine granularity. After the adjustment, the unparsed script text is parsed paragraph by paragraph with the new parsing granularity. This adjustment method effectively solves the problem of incomplete extraction of core information and omission of key details caused by excessively coarse granularity. Granularity of parsing enhances the precision of the analysis, ensuring that all core information is fully extracted. Supported by an extraction frequency-granularity database, granularity adjustments are made more scientific and rational, avoiding biases caused by subjective adjustments. If the frequency of repeated information extraction exceeds a preset upper limit, the current paragraph's parsing granularity is deemed too fine. The current frequency of repeated information extraction is then input into the constructed extraction frequency-granularity database, which outputs an upgrade adjustment command. Specifically, if the current granularity is fine, it is adjusted to medium; if it is medium, it is adjusted to coarse. After adjustment, the unanalyzed script text continues to be parsed segment by segment with the new granularity. This adjustment effectively solves the problems of redundant information extraction, low parsing efficiency, and excessive system resource consumption caused by overly fine granularity. By simplifying the parsing granularity and reducing redundant parsing steps, it focuses on core information extraction, significantly improving the overall efficiency of parsing and information extraction. Simultaneously, the precise matching of the database ensures the rationality of granularity upgrade adjustments, balancing efficiency improvement with the completeness of information extraction, achieving optimal allocation of parsing resources.

[0018] Step 102: The main information of the extracted script text to be processed is matched with the standardized film language database. Based on the matching results, the script text to be processed is decomposed into several independent storyboard units. Based on the time consumption of the storyboard unit decomposition, it is determined whether to perform storyboard unit decomposition precision control. If yes, it is determined whether to perform decomposition efficiency collaborative control after the control. If no, it is determined whether to perform decomposition efficiency collaborative control directly.

[0019] It should be understood that, firstly, the main information automatically extracted from the script text to be processed (including scene information, character action information, core dialogue information, emotional expression information, and key plot points, etc.) is matched in a comprehensive and high-precision manner with a pre-built standardized film language database. The standardized film language database has completed the standardized classification, terminology solidification, and association mapping of core elements related to film storyboards (such as shot type, scene specifications, action corresponding to shot expression, and emotional adaptation to shot language, etc.). Through precise matching, the main information extracted from non-standardized script texts can be transformed into standardized storyboard-related information that conforms to film industry standards. This effectively avoids the deviation in storyboard decomposition caused by non-standardized text expression and messy information format. At the same time, it provides a unified and professional reference for subsequent storyboard unit decomposition, greatly improving the standardization and accuracy of storyboard decomposition and reducing manual correction costs. Subsequently, based on the matching results, the script text to be processed is broken down into several independent storyboard units according to the storyboard logic of standardized film language. Each storyboard unit corresponds to a complete shot expressing a scene, including the core content such as the scene, character actions, dialogue, or emotions under a single shot. This achieves a precise conversion of the script text from textual form to storyboard unit form. This decomposition process relies on database matching results, which can clearly define the boundaries and core elements of storyboard decomposition, avoiding problems such as missing storyboards, over-decomposition, or incomplete decomposition. At the same time, breaking down complex script text into independent small units reduces the complexity of subsequent storyboard processing and provides an operable carrier for subsequent precision and efficiency control. Next, the disassembly time of each storyboard unit is calculated. Based on the disassembly time of each storyboard unit, it is determined whether to perform storyboard unit disassembly precision adjustment. The disassembly time can directly reflect the degree of matching between the current storyboard disassembly precision and disassembly efficiency. If the disassembly time is too long, it may be due to excessive redundancy caused by excessive disassembly precision. If the disassembly time is too short, it may be due to insufficient disassembly precision leading to the loss of core information. By using the disassembly time as a judgment criterion, the precision adjustment can be accurately triggered, avoiding the waste of system resources or the decline in disassembly quality caused by blind adjustment. If it is determined that the precision adjustment of the storyboard unit disassembly needs to be performed, the preset precision adjustment strategy is activated. Based on the core information type, matching degree, and disassembly time of the storyboard unit, the precision standard of the storyboard disassembly is dynamically adjusted (such as refining the disassembly precision of core scene storyboards and simplifying the disassembly precision of secondary plot storyboards). After the adjustment is completed, it is then determined whether to perform disassembly efficiency coordination adjustment. This precision adjustment can achieve dynamic adaptation of the storyboard disassembly precision, which not only ensures the disassembly quality of core storyboard units, but also avoids the inefficiency caused by excessive pursuit of high precision. At the same time, by reducing invalid disassembly operations through precise adjustment, it indirectly improves the overall disassembly efficiency and ensures a balance between storyboard disassembly quality and disassembly precision.If it is determined that no precision adjustment of the storyboard unit is required, then it directly determines whether to perform collaborative adjustment of the dismantling efficiency. This logic avoids ineffective precision adjustment operations, focusing on efficiency optimization and further improving the overall processing speed of storyboard dismantling, provided that the dismantling precision has met the requirements. Whether precision adjustment has been performed or the efficiency judgment stage has been entered directly, the judgment of collaborative adjustment of dismantling efficiency is based on the overall dismantling efficiency of the storyboard unit, the system resource utilization rate, and the dismantling quality. If the overall dismantling efficiency is lower than the preset threshold, the system resource utilization is too high, and the dismantling quality meets the standards, then collaborative adjustment of efficiency is performed to optimize the dismantling process, streamline redundant steps, and achieve a collaborative balance between storyboard dismantling precision and efficiency. If the overall dismantling efficiency meets the preset requirements, no adjustment is needed, and the current dismantling rhythm is maintained to ensure that the storyboard dismantling work is carried out efficiently and with high quality, ultimately achieving the standardization, accuracy, and efficiency of the storyboard dismantling of the script text to be processed.

[0020] It needs to be explained that the specific steps for determining whether to perform storyboard unit disassembly precision adjustment based on the time consumption of storyboard unit disassembly are as follows: The preset segment unit disassembly time interval threshold and redundancy disassembly rate threshold are: the segment unit disassembly time interval threshold includes the upper limit threshold of disassembly time and the lower limit threshold of disassembly time; the redundancy disassembly rate threshold includes the upper limit threshold of redundancy disassembly rate, the lower limit threshold of redundancy disassembly rate, and the target threshold of redundancy disassembly rate. The target threshold for redundancy dismantling rate is within the closed interval formed by the lower limit threshold and the upper limit threshold of redundancy dismantling rate. The upper limit threshold for dismantling time corresponds to the upper limit of triggering redundancy dismantling rate control, and the lower limit threshold for dismantling time corresponds to the lower limit of triggering redundancy dismantling rate control. The system obtains the time taken to disassemble a storyboard unit and the current redundancy rate. The disassembly time is the total time from a single storyboard unit to its formation into an independent storyboard unit. The redundancy rate represents the proportion of storyboard units without narrative significance to the total number of disassembled storyboard units. Storyboard units without narrative significance are those that do not contain any core information such as scene, action, dialogue, or emotion, and cannot be matched with any set of shot language parameters in the standardized film language database.

[0021] The time taken to disassemble the storyboard unit is compared with the preset threshold for the time taken to disassemble the storyboard unit, and the redundancy disassembly rate is dynamically adjusted based on the comparison results.

[0022] The specific steps for dynamically adjusting the redundancy disassembly rate based on the comparison results are as follows: If the disassembly time of the storyboard unit is within the preset disassembly time interval threshold, it is determined that the current disassembly efficiency is within a reasonable range, and the current redundant disassembly rate is kept unchanged. The preset disassembly time interval threshold represents the closed interval formed by the preset lower limit threshold of the storyboard unit disassembly time and the preset upper limit threshold of the storyboard unit disassembly time. If the disassembly time of the storyboard unit is less than the preset lower limit threshold of the disassembly time of the storyboard unit, it is determined that the current disassembly efficiency is too high. The current disassembly time of the storyboard unit is input into the disassembly time-disassembly rate mapping relationship, and the disassembly rate gain is output. The target threshold of the redundant disassembly rate is superimposed with the disassembly rate gain to obtain the target redundant disassembly rate. If the disassembly time of the storyboard unit exceeds the preset upper limit threshold for the disassembly time of the storyboard unit, it is determined that the current disassembly efficiency is too low. The current disassembly time of the storyboard unit is input into the disassembly time-disassembly rate mapping relationship, and the disassembly rate reduction amount is output. The difference between the target threshold for redundant disassembly rate and the disassembly rate reduction amount is processed to obtain the target redundant disassembly rate.

[0023] In this embodiment, preliminary parameter presets are first performed, clearly defining the threshold range for the disassembly time of the storyboard unit and the threshold for the redundant disassembly rate. Specifically, the threshold range for the disassembly time of the storyboard unit includes an upper limit threshold and a lower limit threshold, used to define a reasonable range for disassembly time and provide a clear reference for subsequent disassembly efficiency judgment. The threshold for the redundant disassembly rate includes an upper limit threshold, a lower limit threshold, and a target threshold. The target threshold for the redundant disassembly rate is strictly limited to the closed interval formed by the lower and upper limits. Furthermore, the upper limit threshold for disassembly time corresponds to the upper limit triggering the redundancy rate control, and the lower limit threshold for disassembly time corresponds to the lower limit triggering the redundancy rate control. This preset step provides standardized parameter basis and logical criteria for subsequent disassembly time comparison and dynamic adjustment of the redundant disassembly rate, effectively avoiding the problems of chaotic control and judgment deviations caused by a lack of rules. It ensures the standardization and accuracy of the adjustment process from the source, laying the foundation for achieving a balance between disassembly efficiency and disassembly quality. Subsequently, the time consumed in disassembling storyboard units and the current redundancy rate are accurately obtained. The disassembly time is specifically defined as the total time taken from the start of disassembly to the final formation of an independent and complete storyboard unit, which can intuitively and realistically reflect the disassembly efficiency of a single storyboard unit. The redundancy rate is defined as the proportion of storyboard units without narrative significance to the total number of disassembled storyboard units. It is clearly defined that storyboard units without narrative significance are those that do not contain any core information such as scene, action, dialogue, or emotion, and cannot match any set of shot language parameters in the standardized film language database. This definition accurately distinguishes between effective storyboards and redundant storyboards, avoiding statistical bias in the redundancy rate caused by ambiguity in the determination of redundant storyboards. By accurately obtaining these two core data, the efficiency level and quality of the current storyboard disassembly can be fully grasped, providing reliable data support for subsequent comparison, judgment, and dynamic adjustment, and reducing control errors caused by inaccurate data. Next, the obtained disassembly time of the storyboard unit will be compared with the preset disassembly time interval threshold in a comprehensive and accurate manner. Taking the disassembly time as the core judgment criterion, and combining the logical relationship of the preset threshold, the redundant disassembly rate will be dynamically adjusted to achieve synergistic optimization of disassembly efficiency and disassembly quality. The preset disassembly time interval threshold is a closed interval formed by the preset lower limit threshold and the preset upper limit threshold of the disassembly time of the storyboard unit.The specific adjustment steps and corresponding technical effects are as follows: If the time consumed by the storyboard unit disassembly falls within the preset threshold range of storyboard unit disassembly time, it is directly determined that the current storyboard disassembly efficiency is within a reasonable range. At this time, the current redundancy disassembly rate is kept unchanged. This judgment logic can accurately identify the scenario where the disassembly efficiency is suitable, avoid making ineffective redundancy disassembly rate adjustments, reduce system resource waste, and maintain the current disassembly rhythm and redundancy control standards to ensure that the storyboard disassembly does not have the problem of low efficiency, nor does it generate too many redundant storyboards without narrative meaning, thus ensuring that the disassembly work is carried out efficiently and with high quality. If the disassembly time of a storyboard unit is less than the preset lower limit threshold for storyboard unit disassembly time, it is determined that the current disassembly efficiency is too high. Excessive disassembly efficiency is often accompanied by increased redundant storyboards and decreased disassembly quality. In this case, the current storyboard unit disassembly time is input into a pre-built, well-defined disassembly time-disassembly rate mapping relationship. This mapping relationship accurately outputs the corresponding disassembly rate gain based on the time data. Then, the target threshold for redundant disassembly rate is superimposed on the disassembly rate gain to obtain the target redundant disassembly rate adapted to the current disassembly efficiency. Through this adjustment method, redundant disassembly can be precisely controlled. The improvement in resolution rate avoids excessive redundancy in storyboards due to overly high disassembly efficiency. Furthermore, relying on the mapping relationship between disassembly time and disassembly rate makes the adjustment of the redundancy disassembly rate more scientific and targeted, ensuring that the adjusted redundancy disassembly rate is within a reasonable range, balancing disassembly efficiency and quality. If the disassembly time of a storyboard unit exceeds the preset upper limit threshold for storyboard unit disassembly time, the current disassembly efficiency is determined to be too low. Excessively low disassembly efficiency will lead to overall processing delays and insufficient system resource utilization. In this case, the current storyboard unit disassembly time is also input into the disassembly time-disassembly rate mapping relationship. The mapping relationship outputs the corresponding reduction amount of the disassembly rate. Then, the difference between the target threshold of the redundant disassembly rate and the reduction amount of the disassembly rate is processed to obtain the target redundant disassembly rate adapted to the current scenario. This adjustment can accurately reduce the redundant disassembly rate, reduce the generation of storyboards without narrative significance, thereby simplifying the disassembly process, reducing invalid operations, effectively improving the efficiency of storyboard disassembly, and avoiding the omission of core storyboards due to excessive reduction of the redundant disassembly rate, ensuring that the disassembly quality does not decline. Ultimately, it achieves a dynamic balance between the time consumed in disassembling storyboard units and the redundant disassembly rate, ensuring the efficiency, standardization and high quality of storyboard disassembly work.

[0024] It should be further explained that the specific steps for determining whether to implement coordinated control of dismantling efficiency are as follows: The total number of disassembly threads currently used for storyboard disassembly tasks is obtained in real time. The total number of disassembly threads represents the number of threads that perform storyboard unit disassembly operations. The total number of disassembly threads is compared with the reference range of the total number of threads. Based on the comparison results, the data cache reuse ratio threshold of disassembled storyboard units is dynamically adjusted. The reference range of the total number of threads represents the closed interval formed by the lower reference limit of the total number of threads and the upper reference limit of the total number of threads. If the total number of decomposition threads is within the reference range for the total number of threads, it is determined that the current number of decomposition threads is in the standard load range. The baseline threshold for the reuse ratio of decomposition segmentation unit data cache is maintained to ensure a two-way balance between computational efficiency and data accuracy.

[0025] Determining whether to implement coordinated control of dismantling efficiency also includes: If the total number of disassembly threads is less than the reference lower limit of the total number of threads, it is determined that the current number of disassembly threads is in a low load range. The current total number of disassembly threads is input into the constructed total number of threads-proportion threshold mapping relationship, and the proportion threshold adjustment coefficient is output. The baseline disassembled storyboard unit data cache reuse ratio threshold and the proportion threshold adjustment coefficient are combined to obtain the target disassembled storyboard unit data cache reuse ratio threshold, so as to improve the accuracy of storyboard data disassembly and avoid storyboard detail deviations caused by excessive reuse of cached data. If the total number of disassembly threads exceeds the reference upper limit for the total number of threads, it is determined that the current number of disassembly threads is in a high-load range. The current total number of disassembly threads is input into the established mapping relationship between the total number of threads and the proportional threshold, and the proportional threshold adjustment coefficient is output. The baseline disassembled storyboard unit data cache reuse proportional threshold and the proportional threshold adjustment coefficient are combined to obtain the target disassembled storyboard unit data cache reuse proportional threshold, so as to reduce the computational pressure of the disassembly threads and improve the overall speed of storyboard disassembly.

[0026] In this embodiment, the total number of dismantling threads currently allocated to the storyboard dismantling task is first collected and acquired in real time. This total number of dismantling threads accurately corresponds to the number of threads actually performing storyboard unit dismantling operations, which can intuitively reflect the scale of computing resources currently allocated to the storyboard dismantling task by the system. This provides core computing data support for subsequent load status judgment and control decisions, avoiding control inaccuracies caused by lagging or inaccurate computing data. Subsequently, the real-time acquired total number of dismantling threads is accurately compared with a pre-set reference range for the total number of threads (a closed range consisting of the lower reference limit and the upper reference limit for the total number of threads). Based on the thread load status, the threshold for the data cache reuse ratio of the dismantled storyboard units is dynamically adjusted to achieve accurate matching of computing resources and caching strategies, maximizing the overall efficiency of storyboard dismantling. If the total number of disassembly threads falls within the preset reference range for the total number of threads, it is determined that the current number of disassembly threads is in the standard load range. In this case, the baseline threshold for the data cache reuse ratio of disassembled storyboard units remains unchanged. This strategy accurately matches the cache reuse requirements under standard computing load, avoiding both a decrease in data accuracy due to an excessively high cache reuse ratio and a waste of computing power due to an excessively low reuse ratio, effectively ensuring a two-way balance between storyboard disassembly efficiency and data accuracy. If the total number of disassembly threads is less than the lower limit of the thread total reference range, it is determined that the current number of disassembly threads is in the low load range. In this case, the system has sufficient computing resources but the disassembly efficiency has not reached its upper limit. To further improve the disassembly quality, the current total number of disassembly threads is input into a pre-built mapping relationship between the total number of threads and the ratio threshold. The mapping relationship outputs a corresponding proportional threshold reduction coefficient based on the low-load computing power characteristics. Then, the baseline decomposed storyboard unit data cache reuse ratio threshold is combined with this reduction coefficient to obtain the target cache reuse ratio threshold adapted to low-load scenarios. By reducing the cache reuse ratio, the reliance on historical cached data is reduced, allowing the decomposition calculation to perform more refined analysis based on the original text data, effectively improving the accuracy of storyboard data decomposition and avoiding problems such as storyboard detail deviations and information omissions caused by excessive reuse of cached data. If the total number of decomposition threads exceeds the reference upper limit for the total number of threads, it is determined that the current number of decomposition threads is in a high-load range. At this time, the system's computing resources are strained and the computational pressure is high. To prioritize decomposition efficiency, the current total number of decomposition threads is input to the total number of threads - In the ratio threshold mapping relationship, the corresponding ratio threshold adjustment coefficient is output. Then, the baseline cache reuse ratio threshold is combined with this adjustment coefficient to obtain the target cache reuse ratio threshold adapted to high-load scenarios. By increasing the cache reuse ratio, subsequent storyboard decomposition operations can reuse the data of the already decomposed storyboard units as much as possible, reducing repeated calculation steps, significantly reducing the computational pressure on the decomposition threads, and accelerating the overall speed of storyboard decomposition. At the same time, relying on the accurate support of the total number of threads - ratio threshold mapping relationship, it is ensured that the adjustment range of the cache reuse ratio is reasonable, taking into account both efficiency improvement and data accuracy, and achieving the optimal balance between storyboard decomposition efficiency and quality under different computing power load scenarios.

[0027] Step 103: Each storyboard unit obtained from the decomposition is processed into a structured storyboard data that can be recognized by the machine. The structured storyboard data is then mapped and adapted to the character feature data and scene feature data corresponding to the script text to be processed, so as to generate a visual reference map corresponding to each storyboard unit.

[0028] It should be understood that each independent storyboard unit, after being broken down, undergoes standardized and structured processing. The original text-based storyboard content is converted into structured storyboard data with fixed fields, clear labels, and a standardized format, making it directly readable and recognizable by machines. This eliminates the parsing ambiguity, formatting chaos, and machine recognition obstacles caused by unstructured text, improving the efficiency and stability of subsequent data matching, mapping, and rendering. Then, this standardized structured storyboard data is matched with pre-extracted and solidified character and scene feature data from the script text to be processed, using high-precision character position parameter mapping and adaptation. This is achieved by dynamically matching character poses, character spatial coordinates, character size proportions, and scene layout. Key parameters such as scene perspective relationships and scene lighting conditions are used to achieve three-dimensional collaborative alignment of storyboard semantics, character characteristics, and scene environment, avoiding problems such as character-scene mismatch, spatial relationship confusion, and visual logic contradictions, and greatly improving the rationality and professionalism of visual generation. Finally, based on the completion of precise mapping and adaptation control, corresponding visual reference images that are highly consistent with the content of each storyboard unit, conform to film language norms, and have intuitive reference value are automatically generated. This provides a stable, unified, and directly reusable visual basis for subsequent shot design, screen preview, visual presentation, and production implementation, effectively reducing manual communication costs and repeated modification losses, and improving the overall efficiency of automated conversion from script text to visual storyboard and the quality of the finished product.

[0029] It should be noted that, as Figure 2The diagram shows a flowchart of the character position parameter mapping adaptation and control process for the data decomposition and visualization processing method of the intelligent storyboard provided in this application embodiment. The specific process is as follows: First, the character interaction response time is obtained, and it is determined whether the response time is within a preset interaction response reference range (a closed range consisting of a lower reference limit and an upper reference limit). If it is within the range, the current edge safety distance benchmark value is maintained to balance the integrity of the screen composition and the naturalness of the interaction. If it is not within the range, it is further determined whether the response time is less than the lower reference limit. If so, the response time is input into the response time-safety distance mapping relationship, and the safety distance upward adjustment coefficient is output. The target safety distance is obtained by combining the benchmark value and the upward adjustment coefficient to avoid the character moving out of the frame quickly and to adapt to the real-time requirements of streaming re-rendering. If the response time is greater than the upper reference limit, the response time is input into the response time-safety distance mapping relationship, and the safety distance downward adjustment coefficient is output. The target safety distance is obtained by combining the benchmark value and the downward adjustment coefficient to avoid the character position being too centered, resulting in a hollow screen. Finally, the final safety distance is applied to complete the control.

[0030] It should be further explained that the specific steps for adjusting and adapting character position parameters are as follows: The system collects character interaction commands corresponding to the script text to be processed in real time, and extracts the character interaction response time corresponding to the character interaction command. The character interaction response time represents the time interval from when the character receives the interaction trigger signal to when the character starts the preset interaction action. The interaction trigger signal includes, but is not limited to, the action signals of other characters, the trigger signals of scene elements, and the preset interaction trigger commands in the script text.

[0031] The character's interaction response time is compared with a preset interaction response time threshold range, and the safety distance between the character and the edge of the screen is dynamically adjusted based on the comparison result. Specifically: If the character's interaction response time is within the interaction response reference range, the current edge safety distance baseline value is maintained to balance the integrity of the screen composition and the naturalness of the interaction. The interaction response reference range represents the closed interval formed by the lower limit of the interaction response reference and the upper limit of the interaction response reference range.

[0032] The system dynamically adjusts the safety distance between the character and the edge of the screen based on the comparison results, and also includes: If the character's interaction response time is less than the lower limit of the interaction response reference, the current character's interaction response time is input into the established response time-safe distance mapping relationship, and the safe distance adjustment coefficient is output. The edge safe distance benchmark value and the safe distance adjustment coefficient are combined to obtain the safe distance between the target character and the edge of the screen, so as to avoid the character from exceeding the edge of the screen due to fast movements. At the same time, it adapts to the real-time requirements of the streaming re-rendering architecture and ensures that no screen clipping occurs during the action extension process. If the character's interaction response time is greater than the upper limit of the interaction response reference, the current character's interaction response time is input into the established response time-safe distance mapping relationship, and the safe distance reduction coefficient is output. The edge safe distance benchmark value and the safe distance reduction coefficient are combined to obtain the safe distance between the target character and the edge of the screen, so as to avoid the screen hole caused by the character being too centered.

[0033] In this embodiment, firstly, all character interaction instructions corresponding to the script text to be processed are collected in real time. This comprehensively captures various interactive behavior instructions between characters and between characters and scenes in the storyboard unit, ensuring that no key interactive scenes are missed. This provides complete interactive data support for subsequent response time extraction and spacing adjustment, avoiding one-sided or inaccurate control due to missing interaction instructions. Subsequently, the corresponding character interaction response time is accurately extracted from each collected character interaction instruction. This character interaction response time is specifically defined as the time interval from when the character receives the interaction trigger signal to when the character officially starts the preset interactive action. The scope of the interaction trigger signal is clearly defined, including but not limited to the action signals of other characters, the trigger signals of scene elements, and the preset interaction trigger instructions in the script text. This precise definition can unify the standard for response time extraction, avoid statistical deviations in response time caused by ambiguous trigger signal definitions, and ensure the timeliness of response time data by extracting it in real time. This keeps the data synchronized with the real-time requirements of storyboard rendering and image generation, providing timely and reliable data for subsequent dynamic control. Next, the extracted character interaction response time is compared precisely with a pre-set interaction response time threshold range (i.e., the interaction response reference range, a closed range formed by the lower limit and upper limit of the interaction response reference range). Using response time as the core criterion, the safety distance between the character and the edge of the screen is dynamically adjusted to achieve precise adaptation between the character's interaction actions and the screen composition. This ensures both the integrity of the screen presentation and the naturalness of the character's interaction, enhancing the professionalism and practicality of the visual reference image. The specific adjustment steps and corresponding technical effects are as follows: If the character interaction response time falls within the pre-set interaction response reference range, the current character interaction rhythm is considered appropriate, and no adjustment of the safety distance is needed. The current edge safety distance baseline value remains unchanged. This strategy accurately matches the adaptation requirements of the interaction rhythm and the screen composition, effectively balancing the integrity of the screen composition and the naturalness of the character's interaction. It avoids unnecessary distance adjustments that could lead to screen imbalance or abnormal interaction actions, while also reducing ineffective control operations, lowering system computational pressure, and ensuring efficient screen generation.If the character's interaction response time is less than the lower limit of the interaction response reference, it is determined that the current character's interaction response speed is too fast. If the original safety distance is maintained at this time, the character is very likely to exceed the edge of the screen due to the rapid start of the interaction action, resulting in the problem of action clipping. At the same time, it cannot adapt to the real-time requirements of the streaming re-rendering architecture. Therefore, the current character's interaction response time is input into a pre-built and perfect response time-safety distance mapping relationship. Based on the characteristics of fast response, this mapping relationship accurately outputs the corresponding safety distance adjustment coefficient. Then, the edge safety distance benchmark value and the safety distance adjustment coefficient are scientifically combined and processed to finally obtain the target character and the screen edge safety distance adapted to the current fast interaction scene. This adjustment can effectively expand the safety distance between the character and the screen edge, prevent the character from exceeding the screen edge due to fast action, and perfectly adapt to the real-time requirements of the streaming re-rendering architecture, ensuring that the complete extension process of the character's interaction action does not have any screen clipping problems, and ensuring the continuity and integrity of the screen presentation. If the character's interaction response time exceeds the upper limit of the interaction response reference, it is determined that the current character's interaction response speed is too slow, and the interval between the character's interactive actions is too long. If the original safe distance is maintained at this time, the character's position in the screen will be too centered, resulting in a hollow image and loose composition, which will affect the visual presentation. Therefore, the current character's interaction response time is input into the established response time-safe distance mapping relationship. Based on the characteristics of slow response, this mapping relationship accurately outputs the corresponding safe distance reduction coefficient. Then, the edge safe distance benchmark value and the safe distance reduction coefficient are reasonably combined to obtain the target character and the safe distance between the screen edge that are suitable for the current slow interaction scenario. This adjustment can appropriately reduce the safe distance between the character and the screen edge, optimize the character's position layout in the screen, effectively avoid the problem of a hollow image caused by the character being too centered, make the screen composition more compact and aesthetically pleasing, and at the same time take into account the naturalness of the character's interaction, ensuring that the character's subsequent interactive actions have sufficient screen extension space. Ultimately, a dynamic balance between the character's interaction response rhythm and the safe distance between the screen edge is achieved, which comprehensively improves the image quality and practicality of the storyboard visual reference map and provides a more accurate visual basis for subsequent shot production.

[0034] Step 104: Standardize and verify the visual reference image corresponding to each generated storyboard unit, sort all the visual reference images to form a complete storyboard visualization sequence, complete the automatic conversion from script text to storyboard visual data, and realize the data decomposition and visualization processing of intelligent storyboards.

[0035] It should be understood that, firstly, each visual reference image corresponding to the storyboard unit generated in the early stages undergoes standardized verification. The verification process strictly follows the preset visual specifications, image quality standards, and data format requirements for film storyboards. The focus is on verifying the image clarity, color consistency, character and scene fit, compliance of character position safety spacing, completeness of core interactive actions, and consistency between visual data and structured storyboard data and core information in the script text. At the same time, the format of the visual reference images is verified to meet the requirements for machine recognition and reuse in subsequent production. Through standardized verification, unqualified visual reference images with image distortion, information deviation, format inconsistencies, and abnormal composition can be effectively screened out. This timely avoids interference from unqualified reference images in subsequent production stages, reduces the workload of manual correction, and ensures that each visual reference image is standardized, accurate, and usable. This provides a high-quality visual foundation for the subsequent integration and transformation of storyboard sequences, ensuring the professionalism and reliability of the overall storyboard visualization results. Subsequently, after completing the standardization verification of all visual reference images and confirming their complete compliance, all qualified visual reference images were systematically sorted according to the narrative logic and plot development sequence of the script text to be processed, combined with the order of the storyboard unit breakdown in the early stage. During the sorting process, the plot nodes, character interaction rhythm and scene transition logic of the script text were precisely matched to ensure that the sorted storyboard visualization sequence can completely and coherently restore the core content of the script text, avoiding problems such as disordered storyboard order, broken narrative logic and abrupt scene transitions. Through orderly sorting, a high degree of fit between the visual content of the storyboard and the narrative of the script text is achieved, giving the storyboard visualization sequence a clear narrative thread, intuitively presenting the plot direction and shot logic of the script text, and providing coherent visual guidance for subsequent shot design, screen preview and production implementation. Finally, after completing the standardization and orderly sorting of the visual reference images to form a complete, standardized, and coherent storyboard visualization sequence, the entire process of converting the script text to storyboard visual data can be officially completed. The entire process realizes full-link automation from script text splitting, storyboard decomposition, data control to visual generation, verification, and sorting. It completely eliminates the cumbersome process of manual decomposition, manual drawing, and manual sorting in traditional storyboard production, significantly reducing manual operation costs and human errors, and significantly improving the efficiency and standardization level of storyboard production. Ultimately, it achieves a complete closed loop of intelligent storyboard data decomposition and visualization processing. The generated storyboard visualization sequence can not only intuitively present the visual expression requirements of the script text, but also provide accurate, usable, and reusable storyboard visual data support for subsequent film and television production, animation production, and other stages, promoting the transformation of storyboard production towards intelligence, standardization, and efficiency. At the same time, it ensures that the storyboard visualization results can accurately match the creative intent of the script text, ensuring the consistency between visual presentation and textual expression.

[0036] like Figure 3The diagram shows the structure of the intelligent storyboard data decomposition and visualization processing system provided in this application embodiment. It includes: a script text parsing and extraction efficiency control module, a storyboard unit decomposition and dual-dimensional control module, a storyboard structured processing and visual reference image generation module, and a visual reference image verification, sorting, and complete conversion module. The script text parsing and extraction efficiency control module acquires the script text to be processed, parses it segment by segment based on a pre-built standardized film language database, and determines whether to perform main information extraction efficiency control based on the results of the segmented parsing to improve the efficiency of automatically extracting the main information of the script text. The storyboard unit decomposition and dual-dimensional control module matches the extracted main information of the script text to be processed with the standardized film language database, decomposes the script text to be processed into several independent storyboard units based on the matching results, and then performs dual-dimensional control based on the storyboard units within each storyboard unit. The module determines whether to perform segmentation unit segmentation accuracy adjustment based on the time consumption of segmentation. If so, it determines whether to perform segmentation efficiency coordination adjustment after adjustment; otherwise, it directly determines whether to perform segmentation efficiency coordination adjustment. The segmentation structuring and visual reference map generation module is used to perform structuring processing on each segmentation unit, converting each segmentation unit into machine-recognizable structured segmentation data. It then performs character position parameter mapping and adaptation adjustment on the structured segmentation data and the character feature data and scene feature data corresponding to the script text to be processed, in order to generate a visual reference map corresponding to each segmentation unit. The visual reference map verification, sorting, and complete conversion module is used to perform standardized verification on the visual reference map corresponding to each generated segmentation unit, sort all visual reference maps to form a complete segmentation visualization sequence, complete the automated conversion from script text to segmentation visual data, and realize intelligent segmentation data decomposition and visualization processing.

[0037] In this embodiment, based on the established modules for script text parsing and extraction efficiency control, storyboard unit decomposition and dual-dimensional control, storyboard structuring and visual reference image generation, and visual reference image verification, sorting, and complete conversion, each module executes progressively and interconnects with each other to achieve fully automated processing of the entire process from script text to storyboard visual data. First, the script text parsing and extraction efficiency control module acquires the script text to be processed. Based on a pre-built standardized film language database, it performs standardized parsing of the script text segment by segment. Then, based on the text structure, information density, and language features obtained from the segmented parsing, it determines whether to activate the main information extraction efficiency control. By dynamically optimizing the parsing and extraction strategy, redundant calculations are reduced and the accuracy of key information recognition is improved, thereby significantly improving the efficiency and stability of automatically extracting the core information of the script text, providing a high-quality, high-purity data source for subsequent storyboard processing. Subsequently, the storyboard unit decomposition and dual-dimensional control module performs high-precision matching of the extracted main information of the script text with the standardized film language database. Based on the matched shot language, scene logic, and narrative units, the script text is decomposed into several independent and semantically complete segments. The system analyzes the storyboard units and tracks their disassembly time in real time to determine whether to implement precision control. If precision control is required, it performs precision control first before proceeding to efficiency assessment. If precision control is not required, it directly determines whether to implement collaborative disassembly efficiency control. Through collaborative control of both precision and efficiency, it ensures clear storyboard boundaries and complete content while avoiding over-disassembly or resource waste, achieving optimal quality and processing speed in both aspects. Next, the storyboard structuring and visual reference image generation module performs structuring processing on each disassembled storyboard unit, converting unstructured and non-standard elements into structural elements. The standardized storyboard text content is uniformly converted into structured storyboard data that can be directly recognized, read and processed by machines. Then, the structured storyboard data is matched and adjusted with the character feature data and scene feature data corresponding to the script text through high-precision character position parameter mapping. The character posture, spatial coordinates, lighting and shadow relationship and screen safety distance are dynamically optimized to make semantic information, character features and scene environment highly aligned. Finally, a visual reference map that accurately corresponds to the storyboard unit is automatically generated, effectively eliminating problems such as screen cropping, composition imbalance and character scene mismatch, and improving the standardization, rationality and usability of the visual reference map.Finally, the visual reference image verification, sorting, and complete conversion module performs multi-dimensional standardization checks on all generated visual reference images, including image quality, format specifications, information consistency, and composition rationality. Abnormal data is eliminated, ensuring that each visual reference image is qualified and usable. Then, all visual reference images are uniformly sorted according to the script's narrative order and the storyboard breakdown sequence, forming a logically coherent, narratively complete, and smoothly flowing storyboard visualization sequence. This achieves end-to-end automated conversion from script text to storyboard visual data, minimizing manual intervention, reducing human error, and improving standardization. Ultimately, this completes the closed-loop processing of intelligent storyboards, from text data decomposition to visual presentation.

[0038] Through the above description of the implementation methods, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the above functions can be divided into different functional modules to complete all or part of the functions described above.

[0039] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between devices or units, and may be electrical, mechanical, or other forms.

[0040] The units described as separate components may or may not be physically separate. A component shown as a unit can be one or more physical units, located in one place or distributed in multiple different locations. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0041] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0042] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to the solution, or all or part of the technical solution, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0043] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for data decomposition and visualization processing of intelligent storyboards, characterized in that, Includes the following steps: Step 101: Obtain the script text to be processed. Based on the pre-built standardized film language database, parse the script text to be processed segment by segment. Based on the results of segment-by-segment parsing, determine whether to perform main information extraction efficiency adjustment to improve the efficiency of automatically extracting the main information of the script text to be processed. Step 102: The main information of the extracted script text to be processed is matched with the standardized film language database. Based on the matching results, the script text to be processed is decomposed into several independent storyboard units. Based on the time consumption of the storyboard unit decomposition in the storyboard unit, it is determined whether to perform storyboard unit decomposition precision control. If yes, it is determined whether to perform decomposition efficiency collaborative control after the control. If no, it is determined whether to perform decomposition efficiency collaborative control directly. Step 103: Each storyboard unit obtained from the decomposition is processed into a structured storyboard data that can be recognized by the machine. The structured storyboard data is then mapped and adapted to the character feature data and scene feature data corresponding to the script text to be processed, so as to generate a visual reference map corresponding to each storyboard unit. Step 104: Standardize and verify the visual reference image corresponding to each generated storyboard unit, sort all the visual reference images to form a complete storyboard visualization sequence, complete the automatic conversion from script text to storyboard visual data, and realize the data decomposition and visualization processing of intelligent storyboards.

2. The data decomposition and visualization processing method for intelligent storyboards as described in claim 1, characterized in that: The specific steps for determining whether to perform main information extraction efficiency adjustment based on the results of segment-by-segment parsing are as follows: The system presets a reference range for the frequency of repeated information extraction, presets paragraph parsing granularity levels and corresponding granularity ranges. The paragraph parsing granularity levels include three levels: coarse-grained, medium-grained, and fine-grained. The coarse-grained level corresponds to parsing the smallest text unit as a complete paragraph, the medium-grained level corresponds to parsing the smallest text unit as a sentence, and the fine-grained level corresponds to parsing the smallest text unit as a clause. The frequency of repeated information extraction is obtained. The frequency of repeated information extraction means the number of times the same core information is repeatedly extracted in different parsed text units. The core information is scene information, action information, dialogue information, and emotion information in the script text to be processed. The parsed text unit is the smallest text unit corresponding to the parsing granularity of the current paragraph. The frequency of repeated information extraction is compared with a preset reference range for repeated information extraction, and the granularity of paragraph parsing is dynamically adjusted based on the comparison results. Specifically: If the frequency of repeated information extraction is within the preset reference range for repeated information extraction frequency, it is determined that the parsing granularity of the current paragraph is suitable for the parsing requirements of the current script text. The unparsed script text is then parsed segment by segment using the current parsing granularity. The preset reference range for repeated information extraction frequency represents the closed interval formed by the preset lower limit of the repeated information extraction frequency reference and the preset upper limit of the repeated information extraction frequency reference.

3. The data decomposition and visualization processing method for intelligent storyboards as described in claim 2, characterized in that: The method of dynamically adjusting the granularity of paragraph parsing based on the comparison results also includes: If the frequency of repeated information extraction is less than the preset lower limit of repeated information extraction frequency, it is determined that the current paragraph parsing granularity is too coarse. The current repeated information extraction frequency is input into the constructed extraction frequency-parsing granularity database, and a downgrade adjustment is output. The downgrade adjustment means that if the current granularity is coarse, it is adjusted to medium granularity, and if the current granularity is medium granularity, it is adjusted to fine granularity. After the adjustment, the unparsed script text continues to be parsed paragraph by paragraph with the new parsing granularity. If the frequency of repeated information extraction exceeds the preset upper limit of repeated information extraction frequency, it is determined that the current paragraph parsing granularity is too fine. The current repeated information extraction frequency is input into the constructed extraction frequency-parsing granularity database, and an upgrade adjustment is output. The upgrade adjustment means that if the current granularity is fine, it will be adjusted to medium granularity, and if the current granularity is medium, it will be adjusted to coarse granularity. After the adjustment, the unparsed script text will continue to be parsed paragraph by paragraph with the new parsing granularity.

4. The data decomposition and visualization processing method for intelligent storyboards as described in claim 1, characterized in that: The specific steps for determining whether to perform segmentation unit disassembly accuracy adjustment based on the segmentation unit disassembly time in the segmentation unit are as follows: The preset segmentation unit disassembly time interval threshold and redundancy disassembly rate threshold are defined. The segmentation unit disassembly time interval threshold includes an upper limit threshold for disassembly time and a lower limit threshold for disassembly time. The redundancy disassembly rate threshold includes an upper limit threshold for redundancy disassembly rate, a lower limit threshold for redundancy disassembly rate, and a target threshold for redundancy disassembly rate. The target threshold for the redundancy dismantling rate is within the closed interval formed by the lower threshold for the redundancy dismantling rate and the upper threshold for the redundancy dismantling rate. The upper threshold for dismantling time corresponds to the upper limit of the trigger for redundancy dismantling rate control, and the lower threshold for dismantling time corresponds to the lower limit of the trigger for redundancy dismantling rate control. The time taken to disassemble a storyboard unit and the current redundancy rate are obtained. The time taken to disassemble a storyboard unit is the total time taken from the formation of a single storyboard unit to the formation of an independent storyboard unit. The redundancy rate is the proportion of storyboard units without narrative significance to the total number of disassembled storyboard units. The time taken to disassemble the storyboard unit is compared with the preset threshold for the time taken to disassemble the storyboard unit, and the redundancy disassembly rate is dynamically adjusted based on the comparison results.

5. The data decomposition and visualization processing method for intelligent storyboards as described in claim 4, characterized in that: The specific steps for dynamically adjusting the redundancy disassembly rate based on the comparison results are as follows: If the disassembly time of the storyboard unit is within the preset disassembly time interval threshold, it is determined that the current disassembly efficiency is within a reasonable range, and the current redundant disassembly rate is kept unchanged. The preset disassembly time interval threshold represents the closed interval formed by the lower limit threshold of the preset disassembly time of the storyboard unit and the upper limit threshold of the preset disassembly time of the storyboard unit. If the disassembly time of the storyboard unit is less than the preset lower limit threshold of the disassembly time of the storyboard unit, it is determined that the current disassembly efficiency is too high. The current disassembly time of the storyboard unit is input into the disassembly time-disassembly rate mapping relationship, and the disassembly rate gain is output. The target threshold of the redundant disassembly rate is superimposed with the disassembly rate gain to obtain the target redundant disassembly rate. If the disassembly time of the storyboard unit exceeds the preset upper limit threshold for the disassembly time of the storyboard unit, it is determined that the current disassembly efficiency is too low. The current disassembly time of the storyboard unit is input into the disassembly time-disassembly rate mapping relationship, and the disassembly rate reduction amount is output. The difference between the target threshold for redundant disassembly rate and the disassembly rate reduction amount is processed to obtain the target redundant disassembly rate.

6. The data decomposition and visualization processing method for intelligent storyboards as described in claim 1, characterized in that: The specific steps for determining whether to perform collaborative control of dismantling efficiency are as follows: The total number of disassembly threads currently used for storyboard disassembly tasks is obtained in real time. The total number of disassembly threads represents the number of threads that perform storyboard unit disassembly operations. The total number of disassembly threads is compared with the reference range of the total number of threads. Based on the comparison result, the data cache reuse ratio threshold of disassembled storyboard units is dynamically adjusted. The reference range of the total number of threads represents the closed interval formed by the reference lower limit of the total number of threads and the reference upper limit of the total number of threads. If the total number of decomposition threads is within the reference range for the total number of threads, it is determined that the current number of decomposition threads is in the standard load range. The baseline threshold for the reuse ratio of decomposition segmentation unit data cache is maintained to ensure a two-way balance between computational efficiency and data accuracy.

7. The data decomposition and visualization processing method for intelligent storyboards as described in claim 6, characterized in that: The determination of whether to perform disassembly efficiency coordination control also includes: If the total number of disassembly threads is less than the reference lower limit of the total number of threads, it is determined that the current number of disassembly threads is in a low load range. The current total number of disassembly threads is input into the constructed total number of threads-proportion threshold mapping relationship, and the proportion threshold adjustment coefficient is output. The baseline disassembled storyboard unit data cache reuse ratio threshold and the proportion threshold adjustment coefficient are combined to obtain the target disassembled storyboard unit data cache reuse ratio threshold, so as to improve the accuracy of storyboard data disassembly and avoid storyboard detail deviations caused by excessive reuse of cached data. If the total number of disassembly threads exceeds the reference upper limit for the total number of threads, it is determined that the current number of disassembly threads is in a high-load range. The current total number of disassembly threads is input into the established mapping relationship between the total number of threads and the proportional threshold, and the proportional threshold adjustment coefficient is output. The baseline disassembled storyboard unit data cache reuse proportional threshold and the proportional threshold adjustment coefficient are combined to obtain the target disassembled storyboard unit data cache reuse proportional threshold, so as to reduce the computational pressure of the disassembly threads and improve the overall speed of storyboard disassembly.

8. The data decomposition and visualization processing method for intelligent storyboards as described in claim 1, characterized in that: The specific steps for performing character position parameter mapping adaptation and adjustment are as follows: Real-time acquisition of character interaction instructions corresponding to the script text to be processed, extraction of character interaction response time corresponding to the character interaction instructions, wherein the character interaction response time represents the time interval from when the character receives the interaction trigger signal to when the character starts the preset interaction action; The character's interaction response time is compared with a preset interaction response time threshold range, and the safety distance between the character and the edge of the screen is dynamically adjusted based on the comparison result. Specifically: If the character's interaction response time is within the interaction response reference range, the current edge safety distance baseline value is maintained to balance the integrity of the screen composition and the naturalness of the interaction. The interaction response reference range refers to the closed interval formed by the lower limit of the interaction response reference and the upper limit of the interaction response reference range.

9. The data decomposition and visualization processing method for intelligent storyboards as described in claim 8, characterized in that: The method of dynamically adjusting the safety distance between the character and the edge of the screen based on the comparison results also includes: If the character's interaction response time is less than the lower limit of the interaction response reference, the current character's interaction response time is input into the established response time-safe distance mapping relationship, and the safe distance adjustment coefficient is output. The edge safe distance benchmark value and the safe distance adjustment coefficient are combined to obtain the safe distance between the target character and the edge of the screen, so as to avoid the character from exceeding the edge of the screen due to fast movements. At the same time, it adapts to the real-time requirements of the streaming re-rendering architecture and ensures that no screen clipping occurs during the action extension process. If the character's interaction response time is greater than the upper limit of the interaction response reference, the current character's interaction response time is input into the established response time-safe distance mapping relationship, and the safe distance reduction coefficient is output. The edge safe distance benchmark value and the safe distance reduction coefficient are combined to obtain the safe distance between the target character and the edge of the screen, so as to avoid the screen hole caused by the character being too centered.

10. A system applying the data decomposition and visualization processing method for intelligent storyboards as described in any one of claims 1-9, characterized in that, include: The system includes a script text parsing and extraction efficiency control module, a storyboard unit decomposition and dual-dimensional control module, a storyboard structured processing and visual reference image generation module, and a visual reference image verification, sorting, and complete conversion module. The script text parsing and extraction efficiency control module is used to acquire the script text to be processed, parse the script text to be processed segment by segment based on a pre-built standardized film language database, and determine whether to perform main information extraction efficiency control based on the results of segment-by-segment parsing, so as to improve the efficiency of automatically extracting the main information of the script text to be processed. The storyboard unit decomposition and dual-dimensional control module is used to match the main information of the extracted script text to be processed with a standardized film language database. Based on the matching result, the script text to be processed is decomposed into several independent storyboard units. Based on the time consumption of the storyboard unit decomposition in the storyboard unit, it is determined whether to perform storyboard unit decomposition precision control. If so, it is determined whether to perform decomposition efficiency collaborative control after the control. If not, it is determined directly whether to perform decomposition efficiency collaborative control. The storyboard structuring and visual reference image generation module is used to perform structuring on each storyboard unit obtained from the decomposition, convert each storyboard unit into structured storyboard data that can be recognized by machines, and perform character position parameter mapping and adaptation adjustment on the structured storyboard data and the character feature data and scene feature data corresponding to the script text to be processed, so as to generate a visual reference image corresponding to each storyboard unit. The visual reference image verification, sorting, and complete conversion module is used to standardize and verify the visual reference images corresponding to each generated storyboard unit, sort all the visual reference images to form a complete storyboard visualization sequence, complete the automatic conversion from script text to storyboard visual data, and realize the data decomposition and visualization processing of intelligent storyboards.

Images