Geospatial Object Detection via Synthetic Training Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing automated geospatial object classification systems require labor-intensive software engineering to create trained computer systems for identifying and locating specific objects in geospatial imagery, which reduces efficiency and increases costs, limiting their applicability.
Innovation Solution
A system and method for simplified generation of systems for broad area geospatial object detection, utilizing pre-existing frameworks and machine learning protocols to create synthetic training images and train machine learning classifiers, thereby reducing the need for manual effort and repetitive software engineering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional automated geospatial object classification systems are used, then object detection capability is achieved, but labor-intensive software engineering and high costs are required
Solution Approach 1:
The patent uses synthetic training images (copies/simulations of real images) to train machine learning classifiers, eliminating the need for manual software engineering. The system generates artificial geospatial images with known object locations and characteristics, which serve as training data to automatically teach the classifier how to detect objects, thereby reducing complex manual programming while maintaining detection reliability
Solution Approach 2:
The system enables self-training of classification models through automated synthetic image generation. The machine learning classifier trains itself using the generated synthetic data without requiring extensive manual software engineering or expert intervention, allowing the system to automatically improve its object detection capability through self-service learning
2Reliability
If traditional automated geospatial object classification systems are used, then object detection is performed, but time-consuming manual effort is required
Solution Approach 1:
The patent performs preliminary action by pre-generating synthetic training images with known object characteristics and locations before actual object detection is needed. This pre-computed training data enables the machine learning classifier to be trained in advance, significantly reducing the time required when actual detection tasks need to be performed, as the system already has prepared training materials and pre-trained models
3Measurement precision
If traditional automated geospatial object classification systems are used, then classification accuracy is achieved, but high costs are incurred
Solution Approach 1:
The patent replaces expensive real-world annotated training data with freely generatable synthetic images. By copying and simulating real geospatial scenes with programmatically inserted objects, the system achieves the same training value without the high costs of acquiring, annotating, and managing real annotated imagery, thereby maintaining classification accuracy while dramatically reducing system creation costs
Data Source
AI summary
A system for broad area geospatial object detection includes a processor configured to retrieve training data including a first plurality of orthorectified geospatial training images each including at least one labeled instance of the object of interest, and a second plurality of orthorectified geospatial images each including at least one labeled instance of the object of interest and/or at least one unlabeled instance of the object of interest, and apply at least one type of image correction to the training data. The processor is also configured to train a plurality of machine learning classifier elements, based on the first plurality of orthorectified geospatial training images and subsequently based on the second plurality of orthorectified geospatial images, each of the plurality of machine learning classifier elements being defined by a machine learning protocol parameterized based on one or more visually unique features of the object of interest.