Vehicle Surroundings Imaging with Dynamic Viewpoint Adjustment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing vehicle-mounted camera systems face issues with image compositing, leading to rapid changes in the display image and potential driver confusion due to discontinuities at seam areas where camera viewpoints overlap, resulting in missed obstacles.

Innovation Solution

An image processing apparatus that uses a location obtaining unit, map storage, and state information to transform two-dimensional image data into three-dimensional projection data, adjusting compositing positions and viewpoints based on vehicle state and environment information to create a seamless and recognizable image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Area of stationary object

If multiple cameras are used to composite images of vehicle surroundings, then the field of view is expanded and driver awareness is improved, but discontinuities and seams appear at compositing positions due to viewpoint differences

Engineering Contradiction:
Improvefield of viewVSAvoidimage continuity
Core Design Contradiction:
Area of stationary objectVSReliability

Solution Approach 1:

The patent transforms 2D images from multiple cameras into a 3D space model, then projects them to a new viewpoint. This dimensional transformation allows seamless compositing by mathematically reconciling viewpoint differences in three-dimensional space, eliminating seams and discontinuities while maintaining expanded field of view.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent introduces a 3D space model as an intermediary between the raw camera images and the final composite image. This intermediate representation serves as a common coordinate system that reconciles differences between multiple camera viewpoints, enabling smooth blending without visible seams.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If compositing position is switched based on given conditions, then the display adapts to different driving scenarios, but rapid changes in compositing position cause driver confusion

Engineering Contradiction:
Improvedisplay adaptationVSAvoiddriver recognition
Core Design Contradiction:
Adaptability or versatilityVSEase of operation

Solution Approach 1:

The patent implements dynamic adjustment of the projection viewpoint based on vehicle state information such as steering angle and speed. The viewpoint continuously adapts to match the driver's natural line of sight, providing smooth transitions that maintain driver recognition while adapting to different driving scenarios.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system uses feedback from vehicle state sensors (steering angle, speed, etc.) to continuously adjust the projection viewpoint. This closed-loop control ensures the displayed image naturally follows driver attention, providing adaptability without causing confusion through abrupt changes.

Inventive Principle:
Principle #23Feedback

3Area of stationary object

If images are composited from cameras with overlapping viewing fields, then complete coverage of vehicle surroundings is achieved, but obstacles near seam areas may be missed due to discontinuity

Engineering Contradiction:
Improvesurroundings coverageVSAvoidobstacle detection accuracy
Core Design Contradiction:
Area of stationary objectVSMeasurement precision

Solution Approach 1:

By transforming images into a 3D space model, the patent eliminates seam discontinuities that cause obstacle detection failures. The three-dimensional representation provides continuous spatial coverage, ensuring obstacles near former seam areas are detected without gaps while maintaining complete surroundings coverage.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP2192552B1Image processing apparatus, image processing method, and recording medium
Publication Date: 2018.10.31 FUJITSU LTD
  • EP2192552B1 patent drawingFigure 1
  • EP2192552B1 patent drawingFigure 2
  • EP2192552B1 patent drawingFigure 3

AI summary

An image processing apparatus (10) processing image data obtained by surroundings image of a vehicle through imaging devices (71-74) mounted on the vehicle includes: a state information obtaining unit (11) obtaining state information that indicates a state of driving operation of the vehicle; an image storage unit (5) storing two-dimensional image data pieces obtained from the imaging devices whose viewing fields partly overlap with each other; a specifying unit (12, 13) compositing two-dimensional image data pieces into a shape of a three-dimensional projection plane, and then specifying, in correspondence to the state information, a compositing method for overlapping portions of viewing fields or transformation information to be used for transformation from a three-dimensional image into projection image data; and an image transformation unit (6) transforming the two-dimensional image data pieces into projection image data by using the compositing method or the transformation information specified by the specifying unit (12, 13).