Surround View Image Synthesis with Region-Specific Color Balance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques for surround view systems in automobiles struggle to perform appropriate color balance processing, especially in overlapping regions captured by multiple cameras with different initial color adjustments, leading to inconsistent illumination and discomfort in image synthesis.
Innovation Solution
A semiconductor device with an image acquisition circuit, estimation source image generation circuit, readjustment circuit, and image synthesis circuit that acquires and processes image data from multiple cameras, cancels initial color adjustments, divides images into processing regions, and performs color balance readjustment to ensure suitable color balance across overlapping regions for seamless image synthesis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If color balance processing is performed based on the photographed pixel corresponding to the optical axis, then the processing is simple and fast, but the color balance becomes inappropriate in overlapping regions far from the optical axis
Solution Approach 1:
The patent segments the image into multiple processing regions based on distance from the optical axis. Different color balance processing methods are applied to different segments: central regions use one method while peripheral overlapping regions use another method, allowing optimized processing for each segment's specific requirements
Solution Approach 2:
The patent applies different color balance processing strategies to different local regions of the image. Specifically, overlapping regions far from the optical axis receive specialized processing with larger adjustment gains, while central regions use standard processing, ensuring each region gets the appropriate treatment for its characteristics
2Device complexity
If uniform color balance processing is applied to the entire image, then the processing is simple, but the overlapping regions with different illumination colors cannot be properly adjusted
Solution Approach 1:
The patent introduces dynamic adjustment of processing parameters based on spatial location. The adjustment gain varies dynamically according to the distance from the optical axis and the degree of overlap, allowing the system to adapt processing intensity to local conditions rather than applying a static uniform approach
Solution Approach 2:
The patent changes the processing parameters (adjustment gain) based on the spatial position and overlap characteristics of different regions. By modifying the gain parameter according to distance from the optical axis and overlap degree, the system achieves region-specific color balance adjustment without requiring completely separate processing pipelines
3Manufacturing precision
If different color balance adjustments are applied to overlapping regions, then the color accuracy improves, but the image synthesis becomes more complex and time-consuming
Solution Approach 1:
The patent applies partial action by focusing enhanced color balance processing only on the specific overlapping regions that require it, rather than applying complex processing to the entire image. This selective approach maintains accuracy where needed while minimizing overall processing time
Data Source
AI summary
A semiconductor device includes an image acquisition circuit which acquires a plurality of captured image data obtained by capturing a plurality of images, an estimation source image generation circuit which cancels effects of initial color adjustment processing on each captured image data to generate image data of a plurality of estimation source images, a readjustment circuit which divides each estimation source image into a plurality of processing regions to perform color balance readjustment processing for each processing region, and an image synthesis circuit which synthesizes the image data of the plurality of estimation source images so that overlapping regions included in the estimation source images overlap each other to generate image data of a synthesized image.