No abstract

Indoor magnetic-based positioning has attracted tremendous interests in recent years due to its pervasiveness and independence from extra infrastructure. Existing methods for indoor magnetic-based positioning are either point-based fingerprint matching or sequence-based fingerprint matching using the raw magnetic field strength. However, the magnetometers in smartphones are vulnerable to a few factors such as user's postures and walking speed, which causes the magnetic field strength corresponding to a location often shift in time or exhibit local distortions, thus greatly limits the positioning performance of existing methods rely on raw magnetic field strength. To this end, we observe the differences among magnetic field strength sequences are mainly attributed to small local segments, and design a new sequence-based fingerprint based on the differences among small local segments of raw MFS sequence to represent raw MFS sequence for indoor positioning. To demonstrate the utility of our proposed sequence-based fingerprint, we have performed a comprehensive experimental evaluation on two datasets, the results show that the proposed approach can significantly improve positioning performance compare with baseline methods.

No abstract

Large components docking is an important step in industrial manufacturing. During the docking process, the relative pose of joining components needs real-time and accurate acquisition. Existing methods rely on expensive and complex optical instruments. Photogrammetry has the advantages of low cost and fast measurement speed, but its measurement accuracy decreases sharply along with the increase of the measurement range. This paper proposes a two-stage binocular vision which consists of two sets of binocular vision systems with different accuracy levels in the same coordinate system. Binocular system with corresponding structural parameters is designed for pose measurement at corresponding stages to solve the contradiction between the range and accuracy in measurement. A triangulation and spatial plane fitting method is proposed to calculate relative poses without introducing coordinate transformation errors. We adopt a novel 3D optimization method to further improve the accuracy. Experiment results show that this method can meet the measurement range and accuracy requirements for large components docking. Compared with traditional combined vision measurement station based on multi-vision sensor, the proposed method reduces coordinate transformation errors and overcomes the contradiction between the measurement range and accuracy, which can improve the accuracy and meet the requirements of engineering application.

To discover road anomalies, a large number of detection methods have been proposed. Most of them apply classification techniques by extracting time and frequency features from the acceleration data. Existing methods are time-consuming since these methods perform on the whole datasets. In addition, few of them pay attention to the similarity of the data itself when vehicle passes over the road anomalies. In this article, we propose QF-COTE, a real-time road anomaly detection system via mobile edge computing. Specifically, QF-COTE consists of two phases: (1) Quick filter. This phase is designed to roughly extract road anomaly segments by applying random forest filter and can be performed on the edge node.(2) Road anomaly detection. In this phase, we utilize collective of transformation-based ensembles to detect road anomalies and can be performed on the cloud node. We show that our method performs clearly beyond some existing methods in both detection performance and running time. To support this conclusion, experiments are conducted based on two real-world data sets and the results are statistically analyzed. We also conduct two experiments to explore the influence of velocity and sample rate. We expect to lay the first step to some new thoughts to the field of real-time road anomalies detection in subsequent work.