Orthomosaic Generation via Automated Image Selection and Tonal Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The manual selection of images for orthomosaic generation is time-consuming and costly, and subjective, leading to potential artifacts when combining multiple images from different sources and acquisition times.
Innovation Solution
An automated system that uses a service-oriented architecture with a mosaic generation service, cutline service, tonal balancing service, and image warping service to select and process images based on capture time, generating orthomosaic images with minimal human intervention, utilizing a Hierarchical Mosaic Data Structure (HMDS) to manage and process geospatial images efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual selection of images is used for orthomosaic generation, then the operator can subjectively choose the best source images, but the process becomes time-consuming and costly
Solution Approach 1:
The system enables automated image selection for orthomosaic generation by implementing algorithms that autonomously evaluate and select source images based on quality metrics, eliminating the need for manual human operators to review and choose images, thereby resolving the contradiction between selection quality and processing time
Solution Approach 2:
The patent transforms the subjective human judgment process into an objective automated process by defining and applying specific parameters and algorithms for image quality assessment, allowing the system to automatically select images based on measurable criteria rather than human subjective determination
2Area of stationary object
If multiple separate images are combined into an orthomosaic, then coverage of larger geographic areas is achieved, but artifacts are introduced at image interfaces
Solution Approach 1:
The system applies different processing techniques to different regions of the orthomosaic, specifically implementing tonal balancing and cutline generation at image interfaces to minimize visible differences, while maintaining high resolution and consistency across the entire mosaic by treating boundary regions differently from interior regions
Solution Approach 2:
The patent performs preliminary processing steps including tonal balancing and cutline generation on source images before they are combined into the final orthomosaic, ensuring that interface artifacts are minimized in advance, which allows for seamless stitching and uniform appearance across image boundaries
3Reliability
If the most recent image is preferred for each region, then up-to-date information is provided, but processing complexity increases when managing multiple images from different sources
Solution Approach 1:
The system segments the orthomosaic generation process into distinct functional services including image selection service, cutline service, tonal balancing service, and mosaic generation service, allowing each service to handle specific aspects of managing multiple images from different sources independently, thereby reducing overall system complexity while maintaining the ability to select the most recent images
Data Source
AI summary
A database for creating orthomosaics, where the database is tailored to retrieve the most recent imagery for each pixel, in an efficient manner that partitions the orthomosaic into sufficiently small chunks that can each be quickly processed.