Panorama Imaging Blending Map for Real-Time Stitching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital cameras struggle to stitch high-resolution panoramic images in real-time due to computational intensity, requiring external computers and delaying user access to the final image for printing or sharing.
Innovation Solution
A method for generating panoramic images using a portable imaging device that acquires low-resolution image frames, determines relative displacements, and maps an alpha-blending map from these frames to high-resolution images for seamless stitching, allowing for real-time processing and storage of high-resolution panoramas without user intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution images are stitched using traditional methods, then image quality is improved, but processing time increases dramatically
Solution Approach 1:
The patent segments the stitching process into two distinct phases: first stitching low-resolution preview images to determine alignment and blending parameters, then applying these parameters to stitch the high-resolution main images. This segmentation allows the computationally intensive alignment calculations to be performed on smaller preview images, significantly reducing processing time while maintaining the ability to produce high-quality high-resolution panoramas.
Solution Approach 2:
The patent performs preliminary stitching operations on low-resolution preview images before processing the high-resolution images. The alignment transformations, overlap detection, and blending parameter calculations are all performed in advance using the preview images. This preliminary action establishes all necessary parameters for the final high-resolution stitching, eliminating the need to perform these calculations again on the larger images.
2Measurement precision
If stitching is performed on high-resolution images, then final image quality is improved, but device resource usage increases
Solution Approach 1:
The patent divides the processing workload by separating preview image handling from main image handling. The computationally intensive tasks of feature detection, alignment calculation, and blending parameter determination are performed exclusively on low-resolution preview images. The high-resolution main images only require the simpler task of applying pre-calculated transformations, dramatically reducing the resource burden on embedded device processors and memory systems.
Solution Approach 2:
The patent introduces low-resolution preview images as an intermediary medium between image capture and final panorama production. These preview images serve as proxies for performing all complex computational operations, acting as a mediator that protects the device's limited processing resources from being overwhelmed by direct manipulation of high-resolution images.
3Loss of time
If real-time panorama processing is implemented, then user access time is improved, but computational load increases
Solution Approach 1:
The patent segments processing into immediate preview stitching and deferred main image stitching. The preview images are stitched in real-time using minimal computational resources, providing users with immediate feedback and access to a usable panorama. The high-resolution main images are processed separately with full computational resources when needed, achieving real-time responsiveness without overwhelming the device's processing capabilities.
Data Source
AI summary
A technique of generating a panoramic image involves acquiring a set of at least two main image frames, as well as corresponding generating and/or acquiring corresponding relatively low-resolution image frames, of overlapping portions of a panoramic scene. A map or other information is stored relating to the generation of a low-res panorama. A main panorama image is formed by joining main image frames based on the map or other information gained in the low-res process.