A Sum-of-Squares Polynomial Approach for Road Anomaly Detection Using Vehicle Sensor Measurements

2017 ◽  
Author(s):  
Dule Shu ◽  
Constantino Lagoa ◽  
Timothy Cleary
Keyword(s):  
Feature Extraction ◽  
Anomaly Detection ◽  
Extraction Method ◽  
Sum Of Squares ◽  
New Method ◽  
Feature Points ◽  
Feature Extraction Method ◽  
Different Types ◽  
Simulation Results ◽  
Vehicle Sensor

This paper presents a new method for road anomaly detection. The existence of road anomalies is determined by the behaviors of vehicles. A special polynomial named Sum-of-Squares (SOS) polynomial is used as a metric to evaluate the normality of vehicle behaviors. The method can process multiple types of sensor measurements. A feature extraction method is used to obtain concise representations of the sensor measurements. These representations, called feature points, are used to calculate the value of the SOS polynomial. Simulation results have been shown to demonstrate that the proposed method can effectively detect different types of road anomalies.

scholarly journals A New Skeleton Feature Extraction Method for Terrain Model Using Profile Recognition and Morphological Simplification

2013 ◽  
Vol 2013 ◽  
pp. 1-16
Author(s):  
Huijie Zhang ◽  
Zhiqiang Ma ◽  
Yaxin Liu ◽  
Xinting He ◽  
Yun Ma
Keyword(s):  
Feature Extraction ◽  
Extraction Method ◽  
Feature Points ◽  
Feature Extraction Method ◽  
Terrain Model ◽  
Noise Point ◽  
Key Factor ◽  
Morphological Theory ◽  
Negative Effect ◽  
Real Experimental Data

It is always difficul to reserve rings and main truck lines in the real engineering of feature extraction for terrain model. In this paper, a new skeleton feature extraction method is proposed to solve these problems, which put forward a simplification algorithm based on morphological theory to eliminate the noise points of the target points produced by classical profile recognition. As well all know, noise point is the key factor to influence the accuracy and efficiency of feature extraction. Our method connected the optimized feature points subset after morphological simplification; therefore, the efficiency of ring process and pruning has been improved markedly, and the accuracy has been enhanced without the negative effect of noisy points. An outbranching concept is defined, and the related algorithms are proposed to extract sufficient long trucks, which is capable of being consistent with real terrain skeleton. All of algorithms are conducted on many real experimental data, including GTOPO30 and benchmark data provided by PPA to verify the performance and accuracy of our method. The results showed that our method precedes PPA as a whole.

scholarly journals Analysis on the Travel Characteristics of Sharing Bikes Connected with Different Types of Public Transportation Stations

2021 ◽  
Author(s):  
Xiaohua Yu ◽  
Xiaohui Wang ◽  
Yanna Zhao
Keyword(s):  
Feature Extraction ◽  
Public Transport ◽  
Public Transportation ◽  
Extraction Method ◽  
Supply And Demand ◽  
Feature Extraction Method ◽  
The Public ◽  
Different Types ◽  
Bike Sharing ◽  
Land Types

In order to better solve the problem of unbalanced supply and demand of connected shared bikes, this paper takes shared bikes as the research object, analyzes the usage characteristics of connected bikes in different types of public transport stations, and puts forward a data-based feature extraction method of shared bikes. Firstly, the usage data of shared bikes were collected, and the starting and finishing points were decoded. The public transport stations were divided into five typical types according to the decoded longitude, latitude and surrounding land types. Secondly, the connectivity activity, connectivity distance and user loyalty are put forward as the characteristic indicators of bike-sharing travel. Finally, taking the bicycle data of Chaoyang District of Beijing as an example, the travel characteristic indexes of shared bikes are analyzed. The results show that, as the “last kilometer” travel connecting tool of public transport, the peak of the use of shared bikes connecting residential stations is 6:30 to 9:30, and that of other stations is 7:30 to 9:30. The connecting distance of shared bikes is generally less than 2km, but the connecting distance of office sites can reach 3km, and this site has the highest user loyalty.

A statistical pattern based feature extraction method on system call traces for anomaly detection

2020 ◽  
Vol 126 ◽  
pp. 106348 ◽  
Author(s):  
Zhen Liu ◽  
Nathalie Japkowicz ◽  
Ruoyu Wang ◽  
Yongming Cai ◽  
Deyu Tang ◽  
...  
Keyword(s):  
Feature Extraction ◽  
Anomaly Detection ◽  
Extraction Method ◽  
System Call ◽  
Feature Extraction Method ◽  
Statistical Pattern

scholarly journals Research on Feature Extraction Method of Indoor Visual Positioning Image Based on Area Division of Foreground and Background

2021 ◽  
Vol 10 (6) ◽  
pp. 402
Author(s):  
Ping Zheng ◽  
Danyang Qin ◽  
Bing Han ◽  
Lin Ma ◽  
Teklu Merhawit Berhane
Keyword(s):  
Feature Extraction ◽  
Open Source Software ◽  
Extraction Method ◽  
Feature Matching ◽  
Physical World ◽  
Feature Points ◽  
Feature Extraction Method ◽  
Visual Positioning ◽  
Extraction Algorithm ◽  
Better Than

In the process of indoor visual positioning and navigation, difficult points often exist in corridors, stairwells, and other scenes that contain large areas of white walls, strong consistent background, and sparse feature points. Aiming at the problem of positioning and navigation in the real physical world where the walls with sparse feature points are difficult to be filled with pictures, this paper designs a feature extraction method, ARAC (Adaptive Region Adjustment based on Consistency) using Free and Open-Source Software and tools. It divides the image into foreground and background and extracts their features respectively, to achieve not only retain positioning information but also focus more energy on the foreground area which is favourable for navigation. In the test phase, under the combined conditions of illumination, scale and affine changes, the feature matching maps by the feature extraction algorithm proposed in this paper are compared with those by SIFT and SURF. Experiments show that the number of correctly matched feature pairs obtained by ARAC is better than SIFT and SURF, and whose time of feature extraction and matching is comparable to SURF, which verifies the accuracy and efficiency of the ARAC feature extraction method.

scholarly journals Tracking and Extracting Action Trajectory of Athlete Based on Hierarchical Features

2020 ◽  
Vol 25 (5) ◽  
pp. 677-682
Author(s):  
Tao Pan
Keyword(s):  
Feature Extraction ◽  
Extraction Method ◽  
Poor Performance ◽  
Feature Space ◽  
Image Features ◽  
Feature Extraction Method ◽  
Feature Sets ◽  
Simulation Results ◽  
Hierarchical Features ◽  
Different Levels

The feature extraction from athlete action images is a research hotspot. To accurately evaluate athlete actions, it is necessary to partition the original image into refined blocks, and extract different levels of image features. However, the traditional feature extraction algorithms can only roughly divide action images into several classes, failing to acquire the accurate feature sets of the actions. This leads to relatively poor performance of feature extraction from action images. To overcome the defect of the traditional methods, this paper puts forward a feature extraction method for the action images of badminton players based on hierarchical features. The underlying image features were analyzed based on the techniques of badminton players, and mapped to the feature space of the corresponding dimension. Simulation results show that the proposed method can accurately extract the features from athlete action images.

scholarly journals An Improved Feature Extraction Approach for Web Anomaly Detection Based on Semantic Structure

2021 ◽  
Vol 2021 ◽  
pp. 1-11
Author(s):  
Zishuai Cheng ◽  
Baojiang Cui ◽  
Tao Qi ◽  
Wenchuan Yang ◽  
Junsong Fu
Keyword(s):  
Machine Learning ◽  
Feature Extraction ◽  
Anomaly Detection ◽  
Language Processing ◽  
Web Application ◽  
Extraction Method ◽  
Domain Knowledge ◽  
Semantic Structure ◽  
Feature Extraction Method ◽  
Web Attacks

Anomaly-based Web application firewalls (WAFs) are vital for providing early reactions to novel Web attacks. In recent years, various machine learning, deep learning, and transfer learning-based anomaly detection approaches have been developed to protect against Web attacks. Most of them directly treat the request URL as a general string that consists of letters and roughly use natural language processing (NLP) methods (i.e., Word2Vec and Doc2Vec) or domain knowledge to extract features. In this paper, we proposed an improved feature extraction approach which leveraged the advantage of the semantic structure of URLs. Semantic structure is an inherent interpretative property of the URL that identifies the function and vulnerability of each part in the URL. The evaluations on CSIC-2020 show that our feature extraction method has better performance than conventional feature extraction routine by more than average dramatic 5% improvement in accuracy, recall, and F1-score.

The Improved MFCC Speech Feature Extraction Method and its Application

2013 ◽  
Vol 756-759 ◽  
pp. 4059-4062 ◽  
Author(s):  
Xiao Yan Wang
Keyword(s):  
Feature Extraction ◽  
Extraction Method ◽  
Recognition Rate ◽  
Feature Extraction Method ◽  
Low Snr ◽  
Nonlinear Properties ◽  
Speech Feature ◽  
Simulation Results ◽  
Robust To Noise ◽  
Speech Feature Extraction

Based on traditional MFCC feature, this paper suggests a new kind of speech signal feature: CMFCC by introducing the method of nonlinear properties. Simulation results indicate that the method has a strong robust to noise and is able to enhance the recognition rate under low SNR.

Sign in / Sign up

Export Citation Format

Share Document