Panoramic Image Blending via Raycasted Plane Intersection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing panoramic image blending methods struggle to effectively handle the distortion that occurs during the stitching of omnidirectional images, particularly in three-dimensional spaces, leading to incomplete gradients and challenges in normalizing the blending process.
Innovation Solution
The proposed method employs raycasting to determine overlap regions between images in a three-dimensional space, adjusts the gradient of these overlap regions, and generates a panoramic image by matching the adjusted image planes, thereby preventing edge phenomena and improving blending quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional image stitching methods are used, then the process is simple, but distortion occurs in three-dimensional spaces and gradient completion is incomplete
Solution Approach 1:
The patent introduces raycasting technology to transition from two-dimensional image plane processing to three-dimensional space processing. By casting rays from the camera origin point through image pixels into the three-dimensional scene, the system determines spatial relationships and overlap regions between multiple images, enabling accurate gradient completion and distortion-free stitching in three-dimensional space.
Solution Approach 2:
The patent uses plane intersection points as intermediary elements to bridge multiple image planes. These intersection points, determined through raycasting, serve as reference markers that facilitate gradient calculation and blending between images. The intermediary plane intersection points enable the system to compute accurate gradients without directly processing complex three-dimensional geometry.
2Reliability
If overlap regions are not properly handled, then processing is faster, but edge phenomena occur and blending quality is poor
Solution Approach 1:
The patent performs preliminary raycasting to determine overlap regions and calculate plane intersection points before executing the actual image blending. By pre-computing the spatial relationships and gradient information in the overlap regions, the system avoids edge phenomena and ensures high blending quality while optimizing processing efficiency through structured computation.
3Manufacturing precision
If gradients are not adjusted, then the process is simpler, but distortion and edge phenomena persist
Solution Approach 1:
The patent changes the gradient parameters of image pixels in the overlap regions by calculating corrected gradients based on plane intersection points and raycasting results. This parameter modification adjusts the gradient magnitude and direction to eliminate distortion and edge phenomena, achieving high gradient accuracy through systematic parameter transformation rather than complex processing.
Data Source
AI summary
A panoramic image blending method: includes performing, with reference to the origin, a Raycast on a three-dimensional spatial region including the plurality of captured images; determining whether at least two points of intersection are generated when any one virtual line generated by performing Raycast intersects with planes of the captured images, thereby determining whether there is an overlapping region in which the captured images overlap; generating, on the basis of the overlapping region, a blending region to be synthesized between the captured images; correcting the slopes of a first image plane and a second image plane corresponding to the blending region, with reference to at least one planar point of intersection formed when the first image plane and the second image plane, which are the planes of the captured images, intersect; and generating a panoramic image by matching at least one of the captured images.