De-noising Device for Waterfall Data Trajectory Outlier Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Noise in waterfall data from distributed acoustic sensing systems reduces the accuracy of estimating average vehicle velocity.
Innovation Solution
A de-noising device and method that determines end coordinates of trajectories, extends them to a reference duration, calculates features, detects outlier trajectories, and removes them from the data set.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If noise is present in the waterfall data, then the data contains more information points, but the accuracy of estimating average vehicle velocity deteriorates
Solution Approach 1:
The patent extracts and removes outlier trajectories from the waterfall data that represent noise. By identifying trajectories with abnormal features (such as unusually long durations or inconsistent velocity patterns) and excluding them from the analysis, the system eliminates the harmful noise components while preserving the valid vehicle trajectory data, thereby improving measurement precision without losing useful information.
2Device complexity
If all extracted trajectories are used for velocity calculation, then the calculation is simpler, but the presence of outlier trajectories reduces accuracy
Solution Approach 1:
The patent performs preliminary filtering of trajectories before velocity calculation by detecting and removing outlier trajectories based on feature analysis. This preliminary action identifies and eliminates noisy data points before they can affect the velocity estimation, ensuring that only valid trajectories are used in the final calculation, thus maintaining high accuracy without significantly complicating the overall process.
3Adaptability or versatility
If trajectory duration varies, then more flexible detection is possible, but consistent comparison becomes difficult
Solution Approach 1:
The patent dynamically adjusts the analysis by calculating features for each individual trajectory based on its own characteristics rather than applying fixed uniform criteria. The system evaluates each trajectory's duration, velocity, and other parameters relative to its specific context, allowing flexible handling of varying trajectory lengths while maintaining accurate comparisons through normalized feature extraction and statistical outlier detection.
Data Source
AI summary
To provide a de-noising device, a de-noising method and a computer readable medium capable of reducing the influence of noise included in waterfall data. A de-noising device comprises: a determination unit (241) for determining end coordinates of each of a plurality of trajectories extracted from waterfall data, the waterfall data being acquired by measuring signals from a plurality of sensing portions arranged along a road; an extension unit (242) for extending each trajectory by using the end coordinates so that a duration of time of each 10 trajectory becomes a reference duration of time; a feature calculation unit (243) for calculating a feature of each extended trajectory; a detection unit (244) for detecting an outlier trajectory from the plurality of trajectories based on the features; a removing unit (245) for removing the outlier trajectory from the plurality of trajectories.