Pole Extraction from Optical Imagery Using 3D Triangulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mapping and navigation service providers face challenges in determining the geolocations and attributes of poles and other ubiquitous objects across large geographic areas, as traditional methods are labor-intensive and resource-heavy.
Innovation Solution
A system utilizing machine learning models, particularly deep learning techniques like YOLO and Mask R-CNN, to detect pole-like objects and their semantic keypoints from optical imagery, followed by photogrammetric triangulation to determine three-dimensional coordinates and attributes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional manual methods are used to determine geolocations and attributes of poles, then measurement precision can be maintained, but productivity is significantly reduced and resource requirements increase
Solution Approach 1:
The patent replaces manual mechanical measurement methods with an automated computer-based system that uses machine learning models and photogrammetric triangulation. The system automatically detects pole-like objects in optical imagery, extracts semantic keypoints, performs triangulation to determine 3D coordinates, and generates geolocation data without human intervention, thereby dramatically improving productivity while maintaining precision through algorithmic accuracy.
Solution Approach 2:
The patent introduces an intermediary processing system that acts as a bridge between raw optical imagery and final geolocation data. This intermediary system includes machine learning models for object detection, keypoint extraction modules, and photogrammetric triangulation algorithms that process images to generate accurate pole geolocations, eliminating the need for direct manual measurement while preserving measurement precision.
2Productivity
If automated machine learning methods are used to detect pole-like objects, then productivity is improved, but device complexity increases
Solution Approach 1:
The patent segments the automated pole extraction system into distinct functional modules: (1) machine learning model for pole-like object detection, (2) semantic keypoint extraction module, (3) photogrammetric triangulation module, and (4) 3D coordinate generation module. This segmentation allows each component to be independently developed, tested, and optimized, managing overall system complexity while achieving high productivity through automated processing.
3Measurement precision
If photogrammetric triangulation is performed on multiple redundant observations, then measurement precision is improved, but loss of time increases due to processing multiple images
Solution Approach 1:
The patent performs preliminary action by using machine learning models to detect and identify pole-like objects and their semantic keypoints in multiple images before performing photogrammetric triangulation. This preliminary detection and filtering of relevant features across multiple images allows the triangulation process to focus only on confirmed pole observations, reducing unnecessary processing time while maintaining the precision benefits of using multiple redundant observations for accurate 3D coordinate determination.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables automated and efficient extraction of pole geolocations and attributes, reducing resource requirements and improving accuracy in mapping and navigation applications.
Implementation Method 1
processing a plurality of images using a machine learning model to generate a plurality of redundant observations of a pole-like object (or other object) respectively depicted in the plurality of images
Implementation Method 2
performing a photogrammetric triangulation of the plurality of redundant observations to determine three-dimensional coordinate data of the pole-like object
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An approach is provided for pole extraction from optical imagery. The approach involves, for instance, processing a plurality of images using a machine learning model to generate a plurality of redundant observations of a pole-like object and/or their semantic keypoints respectively depicted in the plurality of images. The approach also involves performing a photogrammetric triangulation of the plurality of redundant observations to determine three-dimensional coordinate data of the pole-like object and/or their semantic keypoints. The approach further involves providing the three-dimensional coordinate data of the pole-like object as an output.