Automated Landmark Visibility Calculation for Navigation Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating landmark-based navigation data are manual, time-consuming, and expensive, making it challenging to keep navigation systems up-to-date with changing road geometries and new landmarks.
Innovation Solution
An automated system using computer vision and deep learning algorithms to identify landmarks and calculate visibility distances from street images, eliminating the need for human intervention and enabling real-time updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual processes are used to identify landmarks and calculate visibility distance, then navigation data can be generated, but the process becomes very time consuming and expensive
Solution Approach 1:
The patent replaces manual mechanical processes with automated computer vision and machine learning systems. Specifically, it uses street view images processed through algorithms to automatically detect landmarks, calculate visibility distances, and generate navigation data, eliminating the need for manual review of street level imagery while maintaining accuracy
Solution Approach 2:
The system enables self-service by allowing the navigation database to automatically update itself using existing street view imagery and geometric data. The automated landmark detection and visibility calculation processes allow the system to generate and update navigation data without human intervention, continuously adapting to road geometry changes and new landmarks
2Reliability
If manual landmark identification is performed, then navigation data is generated, but it becomes challenging to update data when road geometries change or new landmarks are constructed
Solution Approach 1:
The patent implements continuous automated processing of street view images and geometric data to maintain up-to-date navigation information. The system continuously detects changes in road geometries and new landmarks by processing available imagery through machine learning models, ensuring navigation data remains reliable without requiring periodic manual updates
Solution Approach 2:
The system uses feedback loops where generated navigation data is validated against actual street view imagery and geometric information. This feedback mechanism allows the system to automatically detect discrepancies, update landmark positions, and adjust visibility distances when road geometries change or new landmarks appear, maintaining data reliability through continuous verification
3Productivity
If automated computer vision and deep learning algorithms are used, then time and cost are reduced, but hardware requirements and system complexity increase
Solution Approach 1:
The patent employs multi-functional algorithms that perform multiple tasks using the same computational framework. The machine learning models simultaneously detect landmarks, calculate visibility distances, validate navigation data, and adapt to changes, reducing the need for separate specialized systems while maintaining high productivity
Solution Approach 2:
The system uses existing street view imagery and geometric data as templates and reference materials for automated landmark detection. By copying and processing readily available visual information rather than requiring specialized sensing equipment, the system achieves high efficiency without proportionally increasing hardware complexity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An approach is provided for generating navigation data of a geographical location. The approach involves identifying a landmark located along a source road from a source image and segmenting the source image using a deep learning model to identify a segmentation mask. The approach also involves generating a template image based on the segmentation mask and a street image of the landmark, and matching the template image successively with a sequence of images of the landmark to determine a confidence score. The approach further involves, identifying a first image from the sequence of images with confidence score below a predetermined threshold, and selecting a second image with confidence score above the predetermined threshold from the sequence of images. The approach further involves calculating a visibility distance of the landmark based on the source image and the second image, and generating the navigation data based on the calculated visibility distance.