Mobile crowdsensing (MCS) has gained significant attention in recent years and has become an appealing paradigm for urban sensing. For data collection, MCS systems rely on contribution from mobile devices of a large number of participants or a crowd. Smartphones, tablets, and wearable devices are deployed widely and already equipped with a rich set of sensors, making them an excellent source of information. Mobility and intelligence of humans guarantee higher coverage and better context awareness if compared to traditional sensor networks. At the same time, individuals may be reluctant to share data for privacy concerns. For this reason, MCS frameworks are specifically designed to include incentive mechanisms and address privacy concerns. Despite the growing interest in the research community, MCS solutions need a deeper investigation and categorization on many aspects that span from sensing and communication to system management and data storage. In this paper, we take the research on MCS a step further by presenting a survey on existing works in the domain and propose a detailed taxonomy to shed light on the current landscape and classify applications, methodologies, and architectures. Our objective is not only to analyze and consolidate past research but also to outline potential future research directions and synergies with other research areas.

/pdf/a-survey-on-mobile-crowdsensing-systems-challenges-solutions-224oot2vez.pdf

A Survey on Mobile Crowdsensing Systems: Challenges, Solutions, and Opportunities

Microsoft Kinect, a low-cost motion sensing device, enables users to interact with computers or game consoles naturally through gestures and spoken commands without any other peripheral equipment. As such, it has commanded intense interests in research and development on the Kinect technology. In this paper, we present, a comprehensive survey on Kinect applications, and the latest research and development on motion recognition using data captured by the Kinect sensor. On the applications front, we review the applications of the Kinect technology in a variety of areas, including healthcare, education and performing arts, robotics, sign language recognition, retail services, workplace safety training, as well as 3D reconstructions. On the technology front, we provide an overview of the main features of both versions of the Kinect sensor together with the depth sensing technologies used, and review literatures on human motion recognition techniques used in Kinect applications. We provide a classification of motion recognition techniques to highlight the different approaches used in human motion recognition. Furthermore, we compile a list of publicly available Kinect datasets. These datasets are valuable resources for researchers to investigate better methods for human motion recognition and lower-level computer vision tasks such as segmentation, object detection and human pose estimation.

/pdf/a-survey-of-applications-and-human-motion-recognition-with-hags1y076t.pdf

A Survey of Applications and Human Motion Recognition with Microsoft Kinect

The technology of sensor, pervasive computing, and intelligent information processing is widely used in Body Sensor Networks (BSNs), which are a branch of wireless sensor networks (WSNs). BSNs are playing an increasingly important role in the fields of medical treatment, social welfare and sports, and are changing the way humans use computers. Existing surveys have placed emphasis on the concept and architecture of BSNs, signal acquisition, context-aware sensing, and system technology, while this paper will focus on sensor, data fusion, and network communication. And we will introduce the research status of BSNs, the analysis of hotspots, and future development trends, the discussion of major challenges and technical problems facing currently. The typical research projects and practical application of BSNs are introduced as well. BSNs are progressing along the direction of multi-technology integration and intelligence. Although there are still many problems, the future of BSNs is fundamentally promising, profoundly changing the human-machine relationships and improving the quality of people's lives.

https://www.mdpi.com/1424-8220/13/5/5406/pdf

A survey of body sensor networks.

Metrological evaluation of Microsoft Kinect and Asus Xtion sensors

Metrological comparison between Kinect I and Kinect II sensors

In this paper, we present a system using an Android smartphone that collects, displays sensor data on the screen and streams to the central server simultaneously. Bluetooth and wireless Internet connections are used for data transmissions among the devices. Also, using Near Field Communication (NFC) technology, we have constructed a more efficient and convenient mechanism to achieve an automatic Bluetooth connection and application execution. This system is beneficial on body sensor networks (BSN) developed for medical healthcare applications. For demonstration purposes, an accelerometer, a temperature sensor and electrocardiography (ECG) signal data are used to perform the experiments. Raw sensor data are interpreted to either graphical or text notations to be presented on the smartphone and the central server. Furthermore, a Java-based central server application is used to demonstrate communication with the Android system for data storage and analysis.

/pdf/mobile-sensor-data-collector-using-android-smartphone-dd2rrqcmez.pdf

Mobile sensor data collector using Android smartphone

Kinect is a recent technology used for motion detection and human body tracking designed for a video game console. In this study, we explore two different types of 3D image reconstruction methods to achieve a new method for faster and higher quality 3D images. Generating depth perception information using high quality stereo image textures is computationally heavy and inefficient. On the other hand, depth information can be obtained very fast using Kinect but the overall 3D image quality is not refined and it is low resolution. Thus, in this study we explore the combination of higher quality images on a webcam and faster computation of depth information on Kinect in order to create an efficient and enhanced 3D image reconstruction system. This high resolution system has a broad range of applications including 3D motion sensing of human body, hands tracking and finger gestures.

3D image reconstruction and human body tracking using stereo vision and Kinect technology

This paper presents the design of an acoustic sensor array for sound source localization and object detection. An FPGA based data acquisition system is developed for a flexible acoustic sensor array platform. An anechoic chamber is designed to create a clean environment which isolates the experiment from external noises and reverberation echoes. We investigate direction of arrival estimation and source localization experiments with different geometries and with different numbers of sensors. We further present a discussion of parameters that influence the sensitivity and accuracy of the results of these experiments.

3D direction of arrival estimation and localization using ultrasonic sensors in an anechoic chamber

Aiming at the needs of intelligent transportation systems, a lightweight car front detection algorithm is proposed to solve the current YOLOv4 algorithm's complex network structure, more parameters, high requirements for device configuration, slow detection speed, and large memory usage. First of all, this paper replaces the conventional convolution of the feature fusion network with deep separable convolution to reduce the amount of parameters and calculations of the algorithm. Then, the feature map obtained by the backbone extraction network is introduced into the new attention mechanism Coordinate Attention (CA) module to gather effective features. Finally, the Spatial Pyramid Pooling (SPP) module of YOLOv4 is replaced with the SPP-A module that enhances the receptive field to further improve the accuracy of the network's target detection. The experimental results show that the improved algorithm has reduced the model size by 44% compared with the original algorithm, and increased the number of Frames Per Second (FPS) by nearly 10%, but the mean average precision (mAP) is still slightly improved, indicating that the algorithm can meet the requirements of car front detection in actual scenes.

Weidi Jia

Papers

Mobile sensor data collector using Android smartphone

3D image reconstruction and human body tracking using stereo vision and Kinect technology

3D direction of arrival estimation and localization using ultrasonic sensors in an anechoic chamber

Lightweight car front detection algorithm based on improved YOLOv4