Video Style Effect Selection for Accurate Image Conversion

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video processing technologies lack variety and quality in effect props, resulting in suboptimal display effects and content richness, particularly during stylistic image processing.

Innovation Solution

A method and apparatus that determine a stylistic effect type based on image and text, converting them into a target effect image using a video processing apparatus with modules for effect selection, determination, and conversion, enhancing accuracy and display quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If traditional effect props are used in video processing, then the implementation is simple, but the variety and quality of effects are limited

Engineering Contradiction:

Improvevariety of effect propsVSAvoidcomplexity of effect processing system

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system uses a unified style transfer processing mechanism that can handle multiple different effect types (oil painting, watercolor, sketch, etc.) through a single framework. The pre-trained GAN models serve as universal effect generators that can be selected and applied based on different processing needs, eliminating the need for separate specialized modules for each effect type.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system changes the parameter of effect type selection based on input characteristics. By analyzing the input image or video content and determining the appropriate style category, the system dynamically selects which pre-trained GAN model to apply, thereby achieving diverse effects through parameter selection rather than structural complexity.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If stylistic effect processing is performed on images, then creative effects are achieved, but the display effect and image quality are suboptimal

Engineering Contradiction:

Improvestylistic effect capabilityVSAvoidimage quality

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The system performs preliminary classification of the input content to determine the appropriate style category before applying the effect. By pre-training multiple GAN models on different artistic styles and pre-categorizing input content, the system ensures that the most suitable effect model is selected in advance, thereby maintaining high image quality while achieving the desired stylistic transformation.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses pre-trained GAN models that have learned and stored optimal transformation patterns from large datasets of artistic styles. These pre-trained models act as copied knowledge that can be directly applied to new inputs without requiring retraining, ensuring consistent high-quality results across different effect types while reducing computational complexity.

Inventive Principle:

Principle #26Copying

3Adaptability or versatility

If multiple effect types are provided, then user choice increases, but the accuracy of determining the appropriate effect type decreases

Engineering Contradiction:

Improvenumber of effect typesVSAvoidaccuracy of effect type determination

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system segments the effect selection process into distinct style categories (e.g., oil painting, watercolor, sketch, cartoon). By organizing the multiple effect types into well-defined segments with clear characteristics, the system can more accurately determine which category the input belongs to, thereby improving determination accuracy even with a large number of available effects.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary analysis of the input content to identify its characteristics and determine the most appropriate style category before presenting effect options to the user. This preliminary determination step, based on pre-established style categories, ensures accurate effect type selection while maintaining a wide variety of available effects.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20260037115A1Media processing method and device, and storage medium

Publication Date: 2026.02.05 BEIJING ZITIAO NETWORK TECH CO LTD

AI summary

A video processing method includes acquiring effect selection information in response to an effect triggering instruction, determining at least one to-be-applied effect type corresponding to the effect selection information, determining a target effect type from the at least one to-be-applied effect type, and converting the target text and/or the to-be-processed image into a target effect image corresponding to the target effect type. As such, a corresponding stylistic effect type is determined based on an image, text, or a combination thereof; further, the text and the image are converted into the effect of a target effect image based on the corresponding stylistic effect type, improving the accuracy of determining the corresponding stylistic effect type and further improving the image quality and effect display effect of the target effect image.