Vehicle Scenario Image Generation for Realistic Day-to-Night Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image generation methods fail to accurately convert daytime scenario images into night scenario images due to the presence of shadows, resulting in low authenticity of the generated images.
Innovation Solution
An image generation method and apparatus that determines regions in an image where the rendering state does not match the desired scenario information, specifically adjusting the rendering state of vehicles and other objects to match the target scenario, such as adjusting shadows or adding/removing weather elements like rain/snow, using image segmentation networks and light analysis to enhance realism.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If style conversion is performed on daytime scenario images to generate night scenario images, then the scenario transformation is achieved, but shadows remain in the converted images resulting in low authenticity
Solution Approach 1:
The image processing is divided into multiple stages: first performing style conversion to transform the scenario, then separately detecting shadow regions and processing them. This segmentation allows different processing strategies for different parts of the image, resolving the contradiction between scenario transformation and shadow removal.
Solution Approach 2:
Shadow detection and shadow region identification are performed as preliminary actions before final image generation. By identifying shadow regions in advance and processing them separately with appropriate rendering states, the method ensures shadows are removed while maintaining other scenario transformation effects.
2Manufacturing precision
If shadow regions are detected and processed separately, then image authenticity is improved, but processing complexity increases
Solution Approach 1:
A shadow detection module acts as an intermediary between the style conversion module and the final image output. This intermediary identifies shadow regions and directs them for special processing, enabling systematic shadow removal without requiring complete redesign of the image generation pipeline.
Solution Approach 2:
Different rendering states are applied to different regions of the image: shadow regions receive one rendering state (processed to remove shadows) while non-shadow regions receive another rendering state (maintaining scenario transformation). This local differentiation improves authenticity without requiring complete reprocessing of the entire image.
Data Source
AI summary
An image generation method includes: an original image is converted into an image to be processed under a specific scenario; in the image to be processed, a region to be adjusted of which image rendering state does not match scenario information of the specific scenario is determined; the image rendering state of the region to be adjusted is related to light illuminated on a target object in the original image, and the target object includes a vehicle; a target rendering state in the region to be adjusted is determined; the target rendering state matches the scenario information; and in the image to be processed, the image rendering state of the region to be adjusted is adjusted to the target rendering state, to obtain a target image.