Vehicle Camera Image Alignment Using Gravity-Based Pitch and Roll Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distortions occur between augmented reality (AR) images and camera module videos due to tilting or shifting of the camera module post-shipment, particularly in varying loading states, affecting the accuracy of AR navigation.
Innovation Solution
A method and system that utilize a G sensor to determine adjustment values for pitch and roll based on gravity vectors, adjusting video data to align with AR navigation images, and applying offset values to correct distortions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the camera module is installed in the vehicle, then video recording and AR navigation functions are enabled, but distortions occur between AR images and camera videos due to camera tilting or shifting after shipment
Solution Approach 1:
The system performs preliminary calibration by storing reference video data captured when the vehicle is in a known level state (obtaining reference gravity vector). This reference data is used later to compensate for camera module shifts, allowing the system to pre-establish a baseline for correction without requiring physical re-adjustment after shipment.
Solution Approach 2:
The system dynamically adjusts video data by applying rotation matrices calculated from gravity vector changes. When the vehicle's loading state changes, the G sensor detects new gravity vectors, and the processor calculates adjusted rotation matrices to transform the video data, thereby compensating for camera tilting through parameter transformation rather than physical adjustment.
2Measurement precision
If the camera module tolerance adjustment is performed under level-ground condition, then initial calibration is achieved, but the camera posture is affected by vehicle loading states after shipment
Solution Approach 1:
The system continuously monitors gravity vector changes using the G sensor and compares current gravity vectors with the reference gravity vector stored during calibration. Based on this feedback, the processor dynamically recalculates rotation matrices and adjusts video data in real-time, creating a closed-loop system that compensates for posture changes without requiring physical re-calibration.
Solution Approach 2:
The system transitions from a static calibration approach (fixed rotation matrices) to a dynamic adjustment mechanism where rotation matrices are continuously updated based on real-time gravity vector measurements. This allows the system to adapt to changing loading conditions while maintaining alignment accuracy between AR images and camera videos.
3Adaptability or versatility
If the G sensor is used to detect gravity vector changes, then dynamic adjustment for loading states is enabled, but additional sensor and processing requirements are introduced
Solution Approach 1:
The G sensor serves multiple functions: it detects gravity vector changes for video adjustment, determines vehicle stopping conditions, and identifies level ground states for calibration. By making this single sensor multi-functional, the system reduces the need for additional dedicated sensors while achieving loading state compensation.
Solution Approach 2:
The system uses the vehicle's existing G sensor (already present for other vehicle functions) to perform camera alignment compensation, rather than requiring a separate dedicated sensor. The processor also leverages existing video processing capabilities to perform the adjustments, making the system self-sufficient without adding significant external components.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Improves the quality of AR navigation and driving video recording by minimizing distortions caused by camera module position changes due to loading states.
Implementation Method 1
obtaining a gravity vector acquired using a G sensor of the driving video recording device
Data Source
AI summary
A method for adjusting a video image in a vehicle equipped with a driving video recording device, and a vehicle using same, are disclosed. The method may include obtaining, by a camera module of the driving video recording device, video data corresponding to a video depicting surroundings of the vehicle. The method may also include obtaining, by a processor of the driving video recording device, a gravity vector acquired using a G sensor of the driving video recording device. The method may additionally include determining, by the processor, an adjustment value according to the gravity vector acquired using the G sensor. The adjustment value may include one or both of a pitch adjustment value or a roll adjustment value. The method may further include adjusting, by the processor, the video data using the adjustment value to generate adjusted video data. The method may also include displaying, by a head unit of the vehicle, the adjusted video data together with navigation information on a screen.