Vehicle Surround-View Image Stitching With ROI-Based Color Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle camera systems for Advanced Driver Assistance Systems (ADAS) face challenges in combining images from multiple cameras with varying light levels, leading to inconsistencies in brightness and color due to environmental and integration factors, making it difficult to produce a harmonized composite image, especially when cameras have different exposures to ambient light.
Innovation Solution
A system and method that define pre-determined regions of interest for each camera, which are fixed in size and location, allowing for the detection and adjustment of brightness and color before combining images, ensuring a harmonized and constant brightness and color in the composite image, with these regions being defined offline during the vehicle assembly process to minimize computational resources and avoid chassis-related light effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple camera modules are used to capture images from different locations, then a comprehensive top view of the vehicle and surroundings is achieved, but the captured images have different light levels and luminosities due to environmental factors and camera integration differences
Solution Approach 1:
The patent applies local quality by defining specific regions of interest (ROIs) within each camera's captured image. These ROIs are selected to exclude areas affected by chassis reflections and to focus on regions with consistent lighting characteristics. By processing only these specific local regions rather than entire images, the system achieves uniform luminosity across the composite top view while maintaining comprehensive coverage.
2Ease of operation
If cameras are mounted at different locations and positions based on vehicle model, then optimal capture angles are achieved, but the captured images show substantial differences even with similar camera properties
Solution Approach 1:
The patent implements preliminary action by pre-defining the regions of interest (ROIs) for each camera during the vehicle assembly process or before operation. These ROIs are determined offline based on the specific vehicle model and camera mounting positions. By establishing these regions in advance, the system compensates for the inherent differences caused by various camera locations and vehicle models, ensuring consistent image appearance without requiring real-time adjustments.
3Reliability
If dynamic brightness and color adjustment is performed during vehicle operation, then real-time image harmonization is achieved, but computational resources are significantly consumed
Solution Approach 1:
The patent reduces computational complexity by performing the most demanding operations in advance. Specifically, the regions of interest (ROIs) are pre-defined and stored during vehicle assembly or before operation. During real-time operation, the system only needs to extract these pre-defined ROIs and perform simple brightness/color adjustments, rather than dynamically analyzing and defining ROIs for each frame. This preliminary setup significantly reduces the computational burden during vehicle operation.
4Reliability
If the entire captured image is processed for brightness and color adjustment, then complete image harmonization is achieved, but processing time increases
Solution Approach 1:
The patent applies the extraction principle by isolating and processing only the relevant portions of each captured image. Instead of adjusting brightness and color across the entire image, the system extracts specific regions of interest (ROIs) that are most critical for the top view display. These ROIs are carefully selected to exclude areas with chassis reflections and other distracting elements. By processing only these extracted regions, the system achieves effective image harmonization with significantly reduced processing time.
Data Source
AI summary
The system comprises a plurality of cameras, wherein a first camera is configured to acquire a first captured image and a second camera is configured to acquire a second captured image, the first captured image and the second captured image being at least partially overlapped; at least one controller configured to receive and combine at least the first captured image and the second captured image for providing the composite image; and the system is configured to define a first region of interest within the first captured image and a second region of interest within the second captured image; and to detect the at least one of a brightness and a color of at least the first and second region of interest to adjust the at least one of the brightness and the color of at least the first and second captured image before being combined.