Vehicle Surround View Projection Model for Blind Spot Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing around view monitoring (AVM) systems face challenges in allowing drivers to easily confirm the environment around the vehicle, particularly in minimizing blind spots and reducing image distortion, which increases the risk of accidents during parking.
Innovation Solution
An apparatus with imaging devices, sensors, and a controller that divides the vehicle's surroundings into regions, allowing the driver to select a virtual projection model to project and adjust image data, minimizing blind spots and distortion by generating final image data using a virtual imaging device model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If four imaging devices are disposed at front, rear, left and right of the vehicle to provide image data, then the AVM system can capture the environment around the vehicle, but the driver cannot appropriately confirm the environment and blind spots remain
Solution Approach 1:
The patent transitions from a two-dimensional display of four separate camera views to a three-dimensional virtual reality environment that envelops the driver's field of vision. This dimensional transformation allows the system to eliminate blind spots by providing 360-degree spatial awareness, enabling the driver to confirm the environment accurately without traditional blind spots.
Solution Approach 2:
The patent creates a virtual copy of the real-world environment around the vehicle through VR technology. Instead of displaying raw camera feeds, the system generates a replicated three-dimensional representation of the surroundings that the driver can interact with, allowing accurate environmental confirmation from any angle without physical blind spots.
2Manufacturing precision
If traditional AVM systems display image data from multiple cameras, then image data is provided to the driver, but image distortion occurs and blind spots are not minimized
Solution Approach 1:
The patent implements dynamic projection models that automatically adapt to the driver's needs and vehicle state. The system can switch between different projection types (e.g., cylindrical, spherical, planar) based on the selected region or vehicle state changes, optimizing image accuracy for each scenario without requiring complex manual configuration.
Solution Approach 2:
The patent adjusts key projection parameters such as field of view angle, projection distance, and distortion correction factors based on the selected region and vehicle state. By dynamically changing these parameters, the system minimizes image distortion and blind spots while maintaining manageable complexity through automated parameter optimization.
3Ease of operation
If the driver needs to confirm different positions around the vehicle, then comprehensive image data is required, but rapid selection of positions is difficult
Solution Approach 1:
The patent creates a universal virtual reality interface that allows the driver to confirm any position around the vehicle through a single interactive environment. Instead of switching between multiple separate displays or views, the driver can navigate the 360-degree virtual space to access any region of interest, dramatically improving position selection speed and ease of operation.
Solution Approach 2:
The patent implements real-time feedback mechanisms where the system responds to driver interactions (such as head movement or controller input) by dynamically adjusting the virtual camera view. This immediate feedback allows the driver to rapidly explore different positions around the vehicle without delay, reducing the time needed to confirm the environment.
Data Source
Figure 1
Figure 2
Figure 3(a)~3(f)
AI summary
The apparatus includes an input (130, 141, 142) that receives a user input and at least one imaging device (110) that acquires an external image data of the vehicle. A sensor (120) that includes at least one sensor is configured to confirm a state of the vehicle and a display (140) is configured to display a region around the vehicle to be divided into a plurality of regions (1-8) based on the vehicle. A controller (160) selects a virtual projection model based on a selected region or a state change of the vehicle when at least one of the plurality of regions is selected from the user input or the state change of the vehicle is sensed by the sensor, and project the external image data onto the virtual projection model, to generate final image data that corresponds to a position of the selected region.