Feature Triangulation Using Reverse Image Sequence Parallax
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing location-based services and applications, such as autonomous driving, face challenges in achieving high accuracy and detail in digital map data, particularly when features are small relative to an image frame, leading to difficulties in feature identification and triangulation.
Innovation Solution
The system introduces a method to reverse the chronological order of image frames based on vehicle trajectory data, ensuring that features appear larger and closer to the image edges, thereby improving parallax and triangulation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If features are small relative to an image frame, then more geographic area can be covered, but feature identification becomes challenging and triangulation accuracy deteriorates
Solution Approach 1:
The patent inverts the traditional triangulation approach by reversing the chronological order of image frames. Instead of using earlier frames where features are small and near the center, the system uses later frames where features appear larger and closer to image edges, thereby improving feature identification accuracy while maintaining wide geographic coverage
Solution Approach 2:
The system changes the temporal parameter of image frame selection dynamically. By adjusting which frames are used for triangulation based on feature size and position characteristics, the system optimizes measurement precision without sacrificing area coverage
2Difficulty of detecting and measuring
If features are near the image center, then they are easier to detect, but parallax is reduced and triangulation accuracy deteriorates
Solution Approach 1:
The patent reverses the conventional approach by selecting image frames where features are positioned away from the center and closer to edges. This inversion increases parallax effects and improves triangulation accuracy, while the system compensates for detection difficulty through reverse chronological processing
3Device complexity
If traditional chronological triangulation is used, then processing is simpler, but triangulation accuracy deteriorates for small features
Solution Approach 1:
The patent applies reverse chronological processing to select and process image frames in inverted temporal order. This approach improves triangulation accuracy by utilizing frames where features are larger and have better parallax, while maintaining relatively simple processing through systematic frame reordering
Solution Approach 2:
The system performs preliminary sorting and selection of image frames based on reverse chronological order before triangulation processing. This preliminary action ensures that optimal frames are selected in advance, improving accuracy without adding significant complexity to the core triangulation algorithm
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
This approach enhances the accuracy of feature triangulation, reducing errors associated with small feature sizes and proximity to the image center, ultimately leading to improved 3D map data generation.
Implementation Method 1
By reversing the chronological order of a sequence of image frames... the same feature in earlier pairs of image frames of the sequence are usually located at the center with little change between the two images of each pair. In other words, the parallax difference between the same feature in each pair can be very small... This, in turn, can lead to large triangulation errors
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An approach is provided for feature triangulation from images. The approach includes retrieving the plurality of images, wherein the plurality of images is captured by a sensor of a vehicle during a drive; determining a vehicle trajectory of the vehicle during the drive; selecting a first image and a second image from the plurality of images, wherein the first image and second image are arranged in reverse time order based on respective image capture times determined using the vehicle trajectory; after detecting the feature in the first image, processing the second image to detect the feature and to associate the feature as detected in the second image with the feature previously detected in the first image; and processing the detected feature in the first image and the second image to triangulate the location of the feature.