Dual Camera Image Stitching via Adaptive Seam Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Combining images from dual cameras with different focal lengths, such as wide and tele lenses, is challenging due to geometric and photometric misalignments, leading to difficulties in generating a single composite image with improved resolution and detail.
Innovation Solution
A method that aligns wide and tele images to overlap a common field of view, corrects photometric differences, selects a stitching seam, and joins the images to generate a composite image, where one side of the seam is from the wide image and the other side is from the tele image, using techniques like structure guided mesh warping and adaptive alpha blending.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple cameras with different focal lengths are used to capture images, then the field of view coverage and image detail are improved, but the difficulty of aligning and combining the images increases
Solution Approach 1:
The patent introduces an intermediary processing system that includes feature point detection, homography calculation, and warping functions. This intermediary layer mediates between the raw images from different cameras and the final composite image, automatically handling the alignment and integration processes. The system acts as a bridge that transforms multiple images with different perspectives into a unified composite view, resolving the complexity of manual alignment.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting transformation parameters (homography matrices) and blending parameters (alpha values) based on the specific characteristics of each image pair. The system calculates optimal warping parameters for each image to match the target perspective, and adjusts blending weights to smoothly transition between different image regions. This adaptive parameter adjustment resolves the alignment complexity by making the combination process flexible and context-aware.
2Measurement precision
If images from wide and tele lenses are combined, then resolution and detail are improved, but photometric differences and visible artifacts increase
Solution Approach 1:
The patent applies local quality by using adaptive blending where different regions of the composite image have different blending characteristics. The alpha blending function varies spatially, with different weights assigned to wide and tele images in different regions. This allows the system to preserve high-resolution details from the tele image in central regions while using the wide image for peripheral regions, minimizing visible artifacts and photometric differences through location-specific optimization.
Solution Approach 2:
The patent inverts the traditional approach by not simply overlaying one image on another, but rather warping both images to a common reference frame and then blending them bidirectionally. Instead of forcing one image to conform to another, the system transforms both images into a unified coordinate system and uses mutual blending to resolve photometric differences, effectively inverting the unidirectional composition approach and achieving better artifact reduction.
3Adaptability or versatility
If dual camera modules are implemented in electronic devices, then camera functionality is improved, but the difficulty of generating composite images increases
Solution Approach 1:
The patent implements self-service by designing an automated processing pipeline that performs feature detection, matching, homography calculation, warping, and blending without requiring manual intervention. The system automatically detects feature points in each image, calculates the transformation parameters, applies the warping functions, and generates the composite image through adaptive blending. This self-service automation resolves the operational difficulty by making the complex composite image generation process transparent and effortless for users.
Solution Approach 2:
The patent applies preliminary action by pre-calculating and storing homography matrices and transformation parameters for different camera configurations. The system performs preliminary feature point detection and matching to establish transformation relationships before the actual composite image generation. This preliminary processing prepares the data structures and parameters needed for rapid composite image creation, reducing the operational complexity during actual use by having the heavy computational work done in advance.
Data Source
AI summary
A method of generating an image from multiple cameras having different focal lengths is described. The method comprising receiving a wide image and a tele image; aligning the wide image and the tele image to overlap a common field of view; correcting for photometric differences between the wide image and the tele image; selecting a stitching seam for the wide image and the tele image; and joining the wide image and the tele image to generate a composite image, wherein a first portion of the composite image on one side of the stitching seam is from the wide image and a second portion of the composite image on the other side of the stitching seam is from the tele image. An electronic device for generating an image is also described.