Image Processing Apparatus for Seamless Multi-Directional View Synthesis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies fail to effectively combine and display moving images captured from different directions simultaneously, lacking a method to automatically identify and align imaging directions for seamless integration.
Innovation Solution
An image processing apparatus and method that acquires two images with the same timing but different directions, divides one image into partial images based on a predetermined relation with the other, and combines these partial images with the edges of the other image, using motion vectors and vanishing points to align and synthesize the views.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple images from different directions are simultaneously displayed, then the user can see images of views in different directions simultaneously, but the complexity of image alignment and direction matching increases
Solution Approach 1:
The system automatically identifies imaging directions and performs alignment without user intervention. The image processing apparatus autonomously analyzes motion vectors, determines vanishing points, and executes the combining process, making the system self-sufficient in resolving the alignment complexity
Solution Approach 2:
The system transforms images by changing their spatial parameters (position, orientation, scale) based on detected motion vectors and vanishing points. This parameter transformation enables seamless alignment of images from different directions while maintaining realistic perspective relationships
2Loss of information
If images are divided and combined from different directions, then a comprehensive surrounding view is generated, but the processing time and computational load increase
Solution Approach 1:
The system divides images into partial images based on detected vanishing points and motion vector directions. By segmenting the combining process into targeted regions rather than processing entire images, the system achieves comprehensive coverage while reducing unnecessary computational overhead
Solution Approach 2:
The system performs dividing and combining operations only on necessary partial images rather than processing complete images. This partial action approach maintains information completeness while minimizing processing time by focusing computational resources on relevant image segments
3Extent of automation
If automatic direction specification is implemented, then user intervention is eliminated, but the difficulty of detecting and measuring imaging directions increases
Solution Approach 1:
The system replaces manual direction specification with automated computational analysis. By substituting mechanical/user-based direction input with algorithmic analysis of motion vectors and vanishing points, the system achieves automatic alignment while overcoming the difficulty of direct measurement
Solution Approach 2:
The system uses motion vectors and vanishing points as intermediary elements to bridge the gap between raw image data and directional information. These intermediaries facilitate automatic direction detection by providing measurable indicators that represent imaging directions without requiring direct measurement
Data Source
AI summary
From among two images and whose imaging timings are the same and imaging directions are different, one image whose imaging direction has a predetermined relation with that of the other image is specified as a target to be divided, and this target image is divided into a first partial image and a second partial image. Then, the first partial image and the second partial image are combined with the edges of the other image located in the direction perpendicular to the imaging direction, respectively.