Fisheye Image Stitching with Segmented Overlap Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image stitching technologies for panoramic or spherical images struggle with efficient detection of stitching positions, particularly when using multiple wide-angle or fisheye lenses, leading to inefficiencies in processing speed and image quality.
Innovation Solution
An image processing apparatus and method that utilizes two fisheye lenses with orthogonal orientations and different sizes, employing distortion correction and projective transformation to generate a spherical image, with a stitching position detection process that includes pattern matching and an updated conversion table for improved alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If distortion correction and projective transformation are performed on each captured image followed by pattern matching to detect stitching positions, then stitching accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing conversion tables for distortion correction and projective transformation before the actual stitching process. This allows the imaging apparatus to quickly apply pre-computed transformations during real-time stitching operations, significantly reducing processing time while maintaining high stitching position detection accuracy through subsequent pattern matching.
2Measurement precision
If multiple target areas are used for stitching position detection, then stitching accuracy is improved, but device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the overlapping area between captured images into multiple discrete target areas, each with its own conversion table and processing parameters. This allows independent optimization of stitching detection for different regions while managing complexity through modular processing of each segmented area rather than treating the entire overlapping region as a single complex problem.
Data Source
AI summary
An image processing apparatus includes circuitry to input at least a first captured image and a second captured image. The first captured image and the second captured image have an overlapping area. The overlapping area in the first captured image has a plurality of target areas in each of which a stitching position is to be obtained. The plurality of target areas is arranged in an orthogonal direction with respect to a direction in which the first captured image and the second captured image are arranged and have a corresponding plurality of sizes different from each other.