Automated Seamline Construction for High-Resolution Orthomosaics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated systems for constructing high-quality high-resolution orthomosaics from satellite or aerial imagery struggle to create inconspicuous seamlines, especially in complex areas with human development, as prior cost rasters based on tonal similarity are insufficient.
Innovation Solution
A system and method for automated construction of inconspicuous seamlines using a cost raster design that identifies and prioritizes specific image features, such as linear features, to extract seamlines as least cost paths, ensuring seamless transitions across diverse image content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If automated seamline construction uses prior cost rasters based on tonal similarity, then automation is achieved, but seamline quality deteriorates in complex areas with human development
Solution Approach 1:
The patent transforms the cost raster from being based solely on tonal similarity to incorporating multiple parameters including linear feature detection, shadow identification, and development area classification. This multi-parameter approach allows the automated system to distinguish between natural transitions and artificial structures, maintaining high seamline quality in complex developed areas while preserving automation.
Solution Approach 2:
The patent introduces an intermediary cost raster that mediates between the automated processing requirement and the quality requirement. The cost raster acts as a intermediary layer that encodes multiple image characteristics (linear features, shadows, development areas) to guide the seamline extraction algorithm, enabling automated construction to achieve manual-quality results in complex areas.
2Manufacturing precision
If manual construction of seamlines is used, then seamline quality is improved, but labor intensity and time consumption increase
Solution Approach 1:
The patent enables the system to serve itself by automatically detecting and analyzing image features (linear features, shadows, development areas) and using this information to construct high-quality seamlines without manual intervention. The automated cost raster generation and seamline extraction process eliminates the need for manual seamline construction while maintaining quality, significantly reducing time consumption.
3Device complexity
If cost raster is based on tonal similarity only, then processing simplicity is maintained, but seamline inconspicuousness deteriorates in complex areas
Solution Approach 1:
The patent segments the image analysis process into distinct feature detection components: linear feature detection, shadow identification, and development area classification. Each segment processes specific image characteristics independently, then combines them in the cost raster. This segmentation maintains processing simplicity through modular operations while achieving high seamline inconspicuousness by considering multiple feature types.
Data Source
AI summary
A system for semi-automated feature extraction comprising an image analysis server that receives and initializes a plurality of raster images, a feature extraction server that identifies and extracts image features, a mosaic server that assembles mosaics from multiple images, and a rendering engine that provides visual representations of images for review by a human user, and a method for generating a cost raster utilizing the system of the invention.