Animation work infringement detection method based on style feature comparison

Abstract

The invention discloses an animation work infringement detection method based on style feature comparison, and the method comprises the steps: comprehensively extracting a plurality of style features of an animation work to be detected and an original animation work, such as line draft strokes, color matching, composition modes, role design, scene layout, light and shadow effects, texture details and dynamic performance; the method can more accurately reflect the style characteristics of the animation works, improves the accuracy of infringement detection, carries out multi-level preprocessing on the extracted style characteristics, effectively removes noise, enhances key characteristics, provides a more reliable data basis for subsequent quantitative comparison, objectively evaluates the style similarity between the two works through a quantitative comparison formula, and improves the accuracy of infringement detection. According to the method, the influence of subjective factors is reduced, the similarity percentage of the comprehensive style characteristics is compared with the dynamic threshold value, whether infringement suspicion exists in the to-be-detected cartoon works or not is flexibly judged, the content of the output infringement detection report is comprehensive, powerful support is provided for copyright protection, and copyright owners are assisted to better cope with infringement behaviors.

Description

Technical Field

[0001] The present invention relates to the technical field of cartoon work infringement detection, and in particular to a cartoon work infringement detection method based on style feature comparison. Background Art

[0002] With the vigorous development of the animation industry, the number and types of animation works are increasing, and the infringement problem between animation works has gradually become prominent. In recent years, with the continuous advancement of computer vision and image processing technology, the infringement detection method of animation works based on style feature comparison has gradually become a research hotspot. However, the existing infringement detection method of animation works based on style feature comparison still has some shortcomings. On the one hand, the extracted style features are often relatively single and it is difficult to fully reflect the style characteristics of animation works. On the other hand, when comparing the extracted style features, there is a lack of effective preprocessing and quantitative comparison methods, resulting in low accuracy and reliability of the comparison results. Summary of the Invention

[0003] In view of this, the present invention proposes a method for detecting infringement of cartoon works based on style feature comparison, which can effectively solve the defects of the existing technology that the extracted style features are relatively single and the accuracy and reliability of the comparison results are low.

[0004] The technical solution of the present invention is achieved as follows:

[0005] A method for detecting infringement of animation works based on style feature comparison, specifically comprising:

[0006] Extracting multiple style features of the animation work to be tested and the original animation work, wherein the style features include at least five of the following: line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details, and dynamic performance;

[0007] Perform multi-level preprocessing on the extracted style features;

[0008] Perform quantitative comparison on the pre-processed style features to obtain the similarity percentage of the comprehensive style features;

[0009] Compare the similarity percentage of the comprehensive style features with the dynamic threshold to determine whether the animation work under inspection is suspected of infringement;

[0010] Outputting an infringement detection report, which includes similarity percentage, judgment conclusion, detailed style feature comparison data, and infringement risk assessment;

[0011] The quantitative comparison of the pre-processed style features is performed, and the specific comparison formula is:

[0012]

[0013] Among them, S represents the similarity percentage of the comprehensive style features, m represents the number of style feature categories, nj represents the number of style features of the jth category, αj represents the category weight of the jth style feature, wji represents the weight of the i-th feature in the j-th style feature, Fdji represents the i-th feature value in the j-th style feature of the animation work to be detected, and Foji represents the i-th feature value in the j-th style feature of the original animation work.

[0014] As a further optional solution of the method for detecting infringement of cartoon works based on style feature comparison, the category weight αj and feature weight wji are dynamically set according to the importance and discrimination of the style features in the cartoon works, and satisfy and

[0015] As a further optional solution to the method for detecting infringement of animation works based on style feature comparison, the extraction of multiple style features of the animation work to be detected and the original animation work specifically includes:

[0016] Extract line drawing stroke features, including line thickness changes, curvature, continuity, and sharpness of the pen tip;

[0017] Extract color matching features, including main color, color distribution, contrast, and harmony between colors;

[0018] Extract the characteristics of the composition, including the position and layout of the main elements, the setting of the visual focus, and the overall balance of the picture;

[0019] Extract character design features, including the character's outline, clothing details, facial expressions, and body proportions;

[0020] Extract scene layout features, including the scene's perspective, spatial hierarchy, arrangement of background elements, and the degree of integration between the scene and the characters;

[0021] Extract the characteristics of light and shadow effects, including the direction and intensity of the light source, the shape of the shadow, and the creation of light and shadow on the atmosphere of the picture;

[0022] Extract texture detail features, including the texture type on the object surface, the fineness of the texture, and the interaction between texture and light and shadow;

[0023] Extract dynamic performance features, including the smoothness of the character's movements, the rationality of dynamic postures, and the connection between dynamic and static images.

[0024] As a further optional solution to the method for detecting infringement of cartoon works based on style feature comparison, the setting of the dynamic threshold specifically includes:

[0025] Analyze the similarity distribution of past infringement detection cases and determine a reasonable threshold range;

[0026] Set targeted thresholds based on the differences in style characteristics of different types of animation works;

[0027] Based on the influence of the creation era of animation works on the style characteristics, the threshold is adjusted to meet the comparison needs of works from different eras.

[0028] As a further optional solution to the method for detecting infringement of animation works based on style feature comparison, the infringement detection report also includes a visual display of style features with high similarity, and intuitively presents the differences and similarities in style features between the animation work to be detected and the original animation work through comparison charts, heat maps, etc.

[0029] A system for detecting infringement of animation works based on style feature comparison, including:

[0030] A feature extraction module is used to extract multiple style features of the animation work to be tested and the original animation work, wherein the style features include at least five of the following: line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details, and dynamic expression;

[0031] Preprocessing module, used to perform multi-level preprocessing on the extracted style features;

[0032] The quantitative comparison module is used to perform quantitative comparison on the pre-processed style features and calculate the similarity percentage of the comprehensive style features;

[0033] A threshold comparison module is used to compare the similarity percentage of the comprehensive style features with the dynamic threshold to determine whether the animation work to be tested is suspected of infringement;

[0034] A report output module, configured to output an infringement detection report, including a similarity percentage, a judgment conclusion, detailed style feature comparison data, and an infringement risk assessment;

[0035] The quantitative comparison of the pre-processed style features is performed, and the specific comparison formula is:

[0036]

[0037] Among them, S represents the similarity percentage of the comprehensive style features, m represents the number of style feature categories, nj represents the number of style features of the jth category, αj represents the category weight of the jth style feature, wji represents the weight of the i-th feature in the j-th style feature, Fdji represents the i-th feature value in the j-th style feature of the animation work to be detected, and Foji represents the i-th feature value in the j-th style feature of the original animation work.

[0038] A computing device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the processor implements any one of the steps of the above-mentioned method for detecting infringement of animation works based on style feature comparison.

[0039] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for detecting infringement of animation works based on style feature comparison.

[0040] The beneficial effects of the present invention are as follows: by extracting multiple style features of the animation work to be detected and the original animation work, including at least five features such as line drawing strokes, color matching, composition method, character design, scene layout, light and shadow effects, texture details and dynamic performance, this comprehensive feature extraction method can more accurately reflect the style characteristics of the animation work, thereby improving the accuracy of infringement detection, and performing multi-level preprocessing on the extracted style features can remove noise, enhance key features, and further improve the accuracy of subsequent quantitative comparison. Through the quantitative comparison formula, the style features of the animation work to be detected and the original animation work are quantitatively compared to obtain the comprehensive style features. The similarity percentage of the features is a quantitative comparison method that can more objectively evaluate the style similarity between the two works and reduce the influence of subjective factors. By comparing the similarity percentage of the comprehensive style features with the dynamic threshold, it can be determined whether the animation work to be detected is suspected of infringement. The setting of the dynamic threshold can be adjusted according to the actual situation to improve the flexibility and accuracy of the detection. The output infringement detection report includes information such as the similarity percentage, judgment conclusion, detailed style feature comparison data, and infringement risk assessment. This comprehensive report content can provide strong support for copyright protection and help copyright owners better understand the infringement situation and take corresponding measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 This is a flow chart of a method for detecting infringement of cartoon works based on style feature comparison according to the present invention;

[0043] Figure 2 A schematic diagram of the composition of a cartoon work infringement detection system based on style feature comparison according to the present invention;

[0044] Figure 3 A schematic diagram of the composition of a computing device according to the present invention. DETAILED DESCRIPTION

[0045] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] refer to Figures 1 to 3 , a method for detecting infringement of animation works based on style feature comparison, specifically including:

[0047] Extracting multiple style features of the animation work to be tested and the original animation work, wherein the style features include at least five of the following: line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details, and dynamic performance;

[0048] Perform multi-level preprocessing on the extracted style features;

[0049] Perform quantitative comparison on the pre-processed style features to obtain the similarity percentage of the comprehensive style features;

[0050] Compare the similarity percentage of the comprehensive style features with the dynamic threshold to determine whether the animation work under inspection is suspected of infringement;

[0051] Outputting an infringement detection report, which includes similarity percentage, judgment conclusion, detailed style feature comparison data, and infringement risk assessment;

[0052] The quantitative comparison of the pre-processed style features is performed, and the specific comparison formula is:

[0053]

[0054] Among them, S represents the similarity percentage of the comprehensive style features, m represents the number of style feature categories, nj represents the number of style features of the jth category, αj represents the category weight of the jth style feature, wji represents the weight of the i-th feature in the j-th style feature, Fdji represents the i-th feature value in the j-th style feature of the animation work to be detected, and Foji represents the i-th feature value in the j-th style feature of the original animation work.

[0055] In this embodiment, multiple stylistic features are extracted from the animation work under test and the original animation work, including at least five features such as line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details, and dynamic performance. This comprehensive feature extraction method can more accurately reflect the stylistic characteristics of the animation work, thereby improving the accuracy of infringement detection. Multi-level preprocessing of the extracted stylistic features can remove noise, enhance key features, and further improve the accuracy of subsequent quantitative comparison. The stylistic features of the animation work under test and the original animation work are quantitatively compared using a quantitative comparison formula to obtain a similarity percentage of the comprehensive stylistic features. This quantitative comparison method can more objectively assess the stylistic similarity between the two works and reduce the influence of subjective factors. The similarity percentage of the comprehensive stylistic features is compared with a dynamic threshold to determine whether the animation work under test is suspected of infringement. The dynamic threshold setting can be adjusted according to actual circumstances, improving the flexibility and accuracy of detection. The output infringement detection report includes information such as the similarity percentage, judgment conclusion, detailed stylistic feature comparison data, and infringement risk assessment. This comprehensive report content can provide strong support for copyright protection, helping copyright owners better understand infringement situations and take appropriate measures.

[0056] It should be noted that the multi-level preprocessing of the extracted style features specifically includes:

[0057] Preliminary cleaning: remove noise data, irrelevant information, and outliers to ensure the purity and accuracy of style features;

[0058] Normalization: converting style feature values at different scales to the same scale to eliminate dimensionality effects and facilitate subsequent comparative analysis;

[0059] Eigenvalue standardization: Standardize the style eigenvalues to make them conform to the standard normal distribution, improving the stability and reliability of the comparative analysis;

[0060] Feature selection: Based on the importance, relevance, and discriminability of features, the most representative and discriminative style features are selected to reduce feature dimensionality and improve comparison efficiency.

[0061] Feature enhancement: Enhance key style features to highlight their uniqueness and recognizability, further improving the accuracy of infringement detection.

[0062] Preferably, the category weight αj and feature weight wji are dynamically set according to the importance and distinction of the style features in the animation works, and satisfy and

[0063] In this embodiment, the category weight αj and feature weight wji are dynamically set according to the importance and discrimination of the style features in the animation works, so that in the comparison process, more important styles and features can obtain higher weights, thereby improving the accuracy of detection. Different animation works may be more prominent or unique in certain style features. By dynamically adjusting the weights, these differences can be better reflected, making the detection results more in line with the actual situation. The styles of animation works are diverse, and fixed weights may not be able to adapt to all situations. Dynamically setting weights enables the model to flexibly respond to animation works of different styles, improving the generalization ability of detection, and satisfying and The constraints ensure the normalization of weights and avoid the influence of excessive or too small weights on the calculation results, thereby optimizing the calculation efficiency. Dynamic weight adjustment can reduce the calculation of some unimportant features according to actual conditions, thereby improving the overall calculation efficiency. The dynamically set weights can be visualized to help users understand the degree of attention the model pays to different styles and features during the detection process, thereby improving the interpretability of the model. By understanding the weights of different styles and features, users can better understand the detection results and make decisions accordingly.

[0064] Preferably, the extracting of multiple style features of the animation work to be detected and the original animation work specifically includes:

[0065] Extract line drawing stroke features, including line thickness changes, curvature, continuity, and sharpness of the pen tip;

[0066] Extract color matching features, including main color, color distribution, contrast, and harmony between colors;

[0067] Extract the characteristics of the composition, including the position and layout of the main elements, the setting of the visual focus, and the overall balance of the picture;

[0068] Extract character design features, including the character's outline, clothing details, facial expressions, and body proportions;

[0069] Extract scene layout features, including the scene's perspective, spatial hierarchy, arrangement of background elements, and the degree of integration between the scene and the characters;

[0070] Extract the characteristics of light and shadow effects, including the direction and intensity of the light source, the shape of the shadow, and the creation of light and shadow on the atmosphere of the picture;

[0071] Extract texture detail features, including the texture type on the object surface, the fineness of the texture, and the interaction between texture and light and shadow;

[0072] Extract dynamic performance features, including the smoothness of the character's movements, the rationality of dynamic postures, and the connection between dynamic and static images.

[0073] In this embodiment, multiple aspects such as line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details and dynamic performance are covered, ensuring the comprehensive capture of the style characteristics of the animation works. For each style feature, detailed description and extraction are carried out, such as the thickness changes, curvature, continuity of the lines and the sharpness of the pen tip, making the feature extraction more accurate and specific; comprehensive feature extraction enables the model to adapt to animation works of different styles, whether it is traditional hand-painted style or modern digital painting style, and can effectively detect infringement. The detailed feature description and accurate comparison process reduce the possibility of misjudgment and missed judgment, and improve the robustness of the model; the extracted multiple style features provide rich data support for the subsequent comparison process, making the comparison results more credible and convincing. The detailed feature data can help copyright owners or relevant institutions better understand the infringement situation and make decisions accordingly.

[0074] Preferably, the setting of the dynamic threshold specifically includes:

[0075] Analyze the similarity distribution of past infringement detection cases and determine a reasonable threshold range;

[0076] Set targeted thresholds based on the differences in style characteristics of different types of animation works;

[0077] Based on the influence of the creation era of animation works on the style characteristics, the threshold is adjusted to meet the comparison needs of works from different eras.

[0078] In this embodiment, a reasonable threshold range can be determined by analyzing the similarity distribution of past infringement detection cases. This method is based on a large amount of actual data, which makes the threshold setting more scientific and objective, reduces the error caused by subjective judgment, and thus improves the accuracy of detection. Targeted thresholds are set according to the differences in style characteristics of different types of animation works. Different types of animation works (such as hot-blooded, healing, science fiction, etc.) may have significant differences in style characteristics. Targeted threshold settings can better adapt to these differences and improve the accuracy of detection. The style characteristics of animation works will change with the change of the creation year. By setting the threshold based on the creation year of the animation works, the style characteristics of the animation works will change. Adjusting the threshold based on the influence of generations on style features can better adapt to the comparison needs of works from different eras, enhance adaptability and generalization capabilities, and the setting of dynamic thresholds enables the model to flexibly respond to various complex situations. Regardless of whether it is an animation work of different types, styles or eras, accurate infringement detection can be achieved by adjusting the threshold; reasonable threshold settings can more accurately distinguish between infringing and non-infringing works, reduce the possibility of misjudgment (determining non-infringing works as infringing) and missed judgment (determining infringing works as non-infringing), and improve the reliability of detection. By reducing misjudgments and missed judgments, a lot of manual review time can be saved, detection efficiency can be improved, and detection costs can be reduced.

[0079] Preferably, the infringement detection report also includes a visual display of style features with high similarity, and intuitively presents the differences and similarities in style features between the animation work to be detected and the original animation work through comparison charts, heat maps, etc.

[0080] In this embodiment, the differences and similarities in stylistic features between the detected animated work and the original animated work can be intuitively presented through comparison charts, heat maps, and other forms. This visual display makes the detection results easier to understand, allowing even non-professionals to quickly grasp the relationship between the stylistic features of the works. For copyright owners or relevant institutions, the visual display provides more intuitive and comprehensive information, helping them better understand the infringement situation and make decisions accordingly. The visual display not only presents the detection results but also indirectly illustrates part of the detection process, namely how the similarity conclusion is drawn through the comparison of stylistic features. This increases the transparency of the detection and makes users more trustworthy in the detection results. The visual display facilitates user review and verification of the detection results. By viewing the comparison charts and heat maps, users can more intuitively understand which stylistic features lead to high similarity, thereby conducting more in-depth analysis and judgment. The visual display makes technical solutions easier to share and communicate. By displaying comparison charts and heat maps, technicians can more intuitively explain and discuss the detection methods and results, promoting the dissemination and improvement of technology. The visual display also facilitates the collection of user feedback. Users can provide suggestions for improving the detection methods and results by viewing the visualized results, helping technicians continuously optimize and improve the technical solutions.

[0081] A system for detecting infringement of animation works based on style feature comparison, including:

[0082] A feature extraction module is used to extract multiple style features of the animation work to be tested and the original animation work, wherein the style features include at least five of the following: line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details, and dynamic expression;

[0083] Preprocessing module, used to perform multi-level preprocessing on the extracted style features;

[0084] The quantitative comparison module is used to perform quantitative comparison on the pre-processed style features and calculate the similarity percentage of the comprehensive style features;

[0085] A threshold comparison module is used to compare the similarity percentage of the comprehensive style features with the dynamic threshold to determine whether the animation work to be tested is suspected of infringement;

[0086] A report output module, configured to output an infringement detection report, including a similarity percentage, a judgment conclusion, detailed style feature comparison data, and an infringement risk assessment;

[0087] The quantitative comparison of the pre-processed style features is performed, and the specific comparison formula is:

[0088]

[0089] Among them, S represents the similarity percentage of the comprehensive style features, m represents the number of style feature categories, nj represents the number of style features of the jth category, αj represents the category weight of the jth style feature, wji represents the weight of the i-th feature in the j-th style feature, Fdji represents the i-th feature value in the j-th style feature of the animation work to be detected, and Foji represents the i-th feature value in the j-th style feature of the original animation work.

[0090] A computing device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the processor implements any one of the steps of the above-mentioned method for detecting infringement of animation works based on style feature comparison.

[0091] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for detecting infringement of animation works based on style feature comparison.

[0092] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting infringement of animation works based on style feature comparison, characterized in that: Specifically include: Extracting multiple style features of the animation work to be tested and the original animation work, wherein the style features include at least five of the following: line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details, and dynamic performance; Perform multi-level preprocessing on the extracted style features; Perform quantitative comparison on the pre-processed style features to obtain the similarity percentage of the comprehensive style features; Compare the similarity percentage of the comprehensive style features with the dynamic threshold to determine whether the animation work under inspection is suspected of infringement; Outputting an infringement detection report, which includes similarity percentage, judgment conclusion, detailed style feature comparison data, and infringement risk assessment; The quantitative comparison of the pre-processed style features is performed, and the specific comparison formula is: Among them, S represents the similarity percentage of the comprehensive style features, m represents the number of style feature categories, nj represents the number of style features of the jth category, αj represents the category weight of the jth style feature, wji represents the weight of the i-th feature in the j-th style feature, Fdji represents the i-th feature value in the j-th style feature of the animation work to be detected, and Foji represents the i-th feature value in the j-th style feature of the original animation work.

2. The method for detecting infringement of cartoon works based on style feature comparison according to claim 1, characterized in that: The category weight αj and feature weight wji are dynamically set according to the importance and distinction of the style features in the animation works, and meet the requirements of and 3. The method for detecting infringement of cartoon works based on style feature comparison according to claim 2, characterized in that: The extraction of multiple style features of the animation work to be detected and the original animation work specifically includes: Extract line drawing stroke features, including line thickness changes, curvature, continuity, and sharpness of the pen tip; Extract color matching features, including main color, color distribution, contrast, and harmony between colors; Extract the characteristics of the composition, including the position and layout of the main elements, the setting of the visual focus, and the overall balance of the picture; Extract character design features, including the character's outline, clothing details, facial expressions, and body proportions; Extract scene layout features, including the scene's perspective, spatial hierarchy, arrangement of background elements, and the degree of integration between the scene and the characters; Extract the characteristics of light and shadow effects, including the direction and intensity of the light source, the shape of the shadow, and the creation of light and shadow on the atmosphere of the picture; Extract texture detail features, including the texture type on the object surface, the fineness of the texture, and the interaction between texture and light and shadow; Extract dynamic performance features, including the smoothness of the character's movements, the rationality of dynamic postures, and the connection between dynamic and static images.

4. The method for detecting infringement of cartoon works based on style feature comparison according to claim 3, characterized in that: The setting of the dynamic threshold specifically includes: Analyze the similarity distribution of past infringement detection cases and determine a reasonable threshold range; Set targeted thresholds based on the differences in style characteristics of different types of animation works; Based on the influence of the creation era of animation works on the style characteristics, the threshold is adjusted to meet the comparison needs of works from different eras.

5. The method for detecting infringement of cartoon works based on style feature comparison according to claim 4, characterized in that: The infringement detection report also includes a visual display of style features with high similarity, and intuitively presents the differences and similarities in style features between the animation work to be detected and the original animation work through comparison charts, heat maps, etc.

6. A cartoon work infringement detection system based on style feature comparison, characterized by: include: A feature extraction module is used to extract multiple style features of the animation work to be tested and the original animation work, wherein the style features include at least five of the following: line drawing strokes, color matching, composition, character design, scene layout, lighting effects, texture details, and dynamic expression; Preprocessing module, used to perform multi-level preprocessing on the extracted style features; The quantitative comparison module is used to perform quantitative comparison on the pre-processed style features and calculate the similarity percentage of the comprehensive style features; A threshold comparison module is used to compare the similarity percentage of the comprehensive style features with the dynamic threshold to determine whether the animation work to be tested is suspected of infringement; A report output module, configured to output an infringement detection report, including a similarity percentage, a judgment conclusion, detailed style feature comparison data, and an infringement risk assessment; The quantitative comparison of the pre-processed style features is performed, and the specific comparison formula is: Among them, S represents the similarity percentage of the comprehensive style features, m represents the number of style feature categories, nj represents the number of style features of the jth category, αj represents the category weight of the jth style feature, wji represents the weight of the i-th feature in the j-th style feature, Fdji represents the i-th feature value in the j-th style feature of the animation work to be detected, and Foji represents the i-th feature value in the j-th style feature of the original animation work.

7. A computing device, characterized in that The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for detecting infringement of cartoon works based on style feature comparison according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for detecting infringement of animation works based on style feature comparison as recited in any one of claims 1 to 5.

Images