Smartphones have become an indispensable tool in the work and life of most people. Given the many kinds of sensors built into smartphones, solving indoor navigation problems using these phones has become feasible. This paper proposes such a solution, using a particle-filter-based indoor positioning method. This method improves the positioning accuracy and helps improve the user experience. The major features of the method are that it leverages and fuses the advantages of two different indoor positioning methods [pedestrian dead reckoning (PDR) and received-signal-strength-indicator-based indoor positioning], and considers the influence on positioning of the indoor spatial map. The smartphone built-in sensors are used as data sources and constitute a feasible method for PDR. The iBeacon devices are also used--because of the many advantages, they present in indoor environments, such as low energy consumption, convenient deployment, and popular applications--and their received signal strength indicator values are used as observations in the particle filter. The devices used in this paper are conventional, without any special equipment, and the experimental sites are common work environments. Therefore, we believe that the proposed method has a wide application potential for indoor navigation applications (e.g., museum guides), and is especially suitable for working with indoor navigation maps.

It has traditionally been laborious and problematic to construct and maintain fingerprint maps for indoor positioning. In this paper, a multi-phase fingerprint map indoor localization method based on interpolation was proposed. The method addresses two needs: (1) to construct an efficient fingerprint map, and (2) to provide high-accuracy indoor positioning. For the effective collection of the real-time received signal strength indicator (RSSI), we proposed a mobile data-collection cart, equipped with a laser rangefinder and a smartphone, which can rapidly collect environmental data in indoor spaces. In practice, as some regions might be inaccessible, we adopted a kriging-based interpolation method that exploits the spatial autocorrelation of the RSSI to efficiently generate and update the fingerprint database. To overcome the instability of RSSI and improve positioning accuracy, we proposed a time-variant multi-phase fingerprint map, with specific fingerprint databases constructed for different time periods, thereby automatically employing the most appropriate fingerprint map according to the time period. Indoor experiments verified that the proposed method has practical value, with sufficient accuracy for general indoor applications. We believe that the proposed method provides a feasible and low-cost indoor positioning solution.

Noise mapping is an effective method of visualizing and accessing noise pollution. In this paper, a noise-mapping method based on smartphones to effectively and easily measure environmental noise is proposed. By using this method, a noise map of an entire area can be created using limited measurement data. To achieve the measurement with certain precision, a set of methods was designed to calibrate the smartphones. Measuring noise with mobile phones is different from the traditional static observations. The users may be moving at any time. Therefore, a method of attaching an additional microphone with a windscreen is proposed to reduce the wind effect. However, covering an entire area is impossible. Therefore, an interpolation method is needed to achieve full coverage of the area. To reduce the influence of spatial heterogeneity and improve the precision of noise mapping, a region-based noise-mapping method is proposed in this paper, which is based on the distribution of noise in different region types tagged by volunteers, to interpolate and combine them to create a noise map. To validate the effect of the method, a comparison of the interpolation results was made to analyse our method and the ordinary Kriging method. The result shows that our method is more accurate in reflecting the local distribution of noise and has better interpolation precision. We believe that the proposed noise-mapping method is a feasible and low-cost noise-mapping solution.

Land cover is an important variable of the terrestrial ecosystem that provides information for natural resources management, urban sprawl detection, and environment research. To classify land cover with high-spatial-resolution multispectral remote sensing imagery is a difficult problem due to heterogeneous spectral values of the same object on the ground. Fully convolutional networks (FCNs) are a state-of-the-art method that has been increasingly used in image segmentation and classification. However, a systematic quantitative comparison of FCNs on high-spatial-multispectral remote imagery was not yet performed. In this paper, we adopted the three FCNs (FCN-8s, Segnet, and Unet) for Gaofen-2 (GF2) satellite imagery classification. Two scenes of GF2 with a total of 3329 polygon samples were used in the study area and a systematic quantitative comparison of FCNs was conducted with red, green, blue (RGB) and RGB+near infrared (NIR) inputs for GF2 satellite imagery. The results showed that: (1) The FCN methods perform well in land cover classification with GF2 imagery, and yet, different FCNs architectures exhibited different results in mapping accuracy. The FCN-8s model performed best among the Segnet and Unet architectures due to the multiscale feature channels in the upsampling stage. Averaged across the models, the overall accuracy (OA) and Kappa coefficient (Kappa) were 5% and 0.06 higher, respectively, in FCN-8s when compared with the other two models. (2) High-spatial-resolution remote sensing imagery with RGB+NIR bands performed better than RGB input at mapping land cover, and yet the advantage was limited; the OA and Kappa only increased an average of 0.4% and 0.01 in the RGB+NIR bands. (3) The GF2 imagery provided an encouraging result in estimating land cover based on the FCN-8s method, which can be exploited for large-scale land cover mapping in the future.

Simultaneous localization and mapping (SLAM) systems have been generally limited to static environments. Moving objects considerably reduce the location accuracy of SLAM systems, rendering them unsuitable for several applications. Using a combined vision camera and inertial measurement unit (IMU) to separate moving and static objects in dynamic scenes, we improve the location accuracy and adaptability of SLAM systems in these scenes. We develop a moving object-matched feature points elimination algorithm that uses IMU data to eliminate matches on moving objects but retains them on stationary objects. Moreover, we develop a second algorithm to validate the IMU data to avoid erroneous data from influencing image feature points matching. We test the new algorithms with public datasets and in a real-world experiment. In terms of the root mean square error of the location absolute pose error, the proposed method exhibited higher positioning accuracy for the public datasets than the traditional algorithms. Compared with the closed-loop errors obtained by OKVIS-mono and VINS-mono, those obtained in the practical experiment were lower by 50.17% and 56.91%, respectively. Thus, the proposed method eliminates the matching points on moving objects effectively and achieves feature point matching results that are realistic.

High-precision indoor positioning is important for modern society. This paper proposes a way to achieve high positioning accuracy and obtain a trajectory close to the actual path in a common application scenario by smartphone without the use of a complicated algorithm. In the actual positioning process, a stable signal source can reduce the signal interference caused by environments. Bluetooth low energy has its own advantages in indoor positioning because it can be seen as a more stable signal source. In this study, we used smartphones to record the changing Bluetooth signals and used a basic nearest neighbor, weight centroid, and probability-based method, which we called an advanced weighted centroid method, to obtain position coordinates and the motion trajectory during the experiment. We used a weight centroid method based on least squares to solve the overdetermined problem. This can also be used to calculate the initial position of the advanced weight centroid. The advanced weighted centroid method introduced a Gaussian distribution to model the distribution of the signal. Translating a deterministic problem into a fuzzy probability problem aligns more with positioning facts and can achieve better results. Experimental results showed that the root-mean-square error (RMSE) of the dynamic positioning result obtained through the probabilistic method was within 1 m and had a more consistent trajectory. Moreover, the impact of the number of iBeacons on the positioning accuracy has been discussed, and a reference for iBeacon placement has been provided. In addition, an experiment was also conducted on the effect of signal transmission frequency on accuracy.