Entity

Time filter

Source Type


Mei L.,CAS Institute of Remote Sensing Applications | Mei L.,University of Chinese Academy of Sciences | Xue Y.,CAS Institute of Remote Sensing Applications | Xue Y.,London Metropolitan University | And 11 more authors.
International Journal of Remote Sensing | Year: 2012

Aerosol optical thickness (AOT) retrieval from Moderate Resolution Imaging Spectroradiometer (MODIS) data has been well established over oceans, but this is not the case over land. In this article, the AOT data sets retrieved by exploiting the synergy of TERRA and AQUA MODIS data (SYNTAM) over land are validated with ground-based measurements from Aerosol Robotic Network (AERONET) data, as well as from the National Aeronautics and Space Administration (NASA) AOT products, amended with a DeepBlue algorithm in Asian (15-60° N and 35-150° E) and American areas (30-40° N and 100-120° W). Overall, AOT retrieval errors of around 10-20% against AERONET data are found at both 1 and 10 km resolutions. The spectral and spatial sensitivities of the AOT correlation are explicitly addressed at both 1 and 10 km resolutions. Three window sizes, 1 × 1, 3 × 3 and 5 × 5, are tested for SYNTAM to evaluate the effect of window size on parameter statistics, and it is found that the accuracy of the SYNTAM method decreases with increasing window size. The validations at three spectral bands of 0.47, 0.55 and 0.66 μm show that the accuracies of different bands are 80-90% similar, and that the band at 0.47 μm has the highest accuracy most of the time. Comparisons between AOT data sets derived from the SYNTAM and AOT products from the NASA Dark Dense Vegetation (DDV) and the DeepBlue algorithms are also conducted using data from the USA. More pixels with AOT values for the area could be retrieved using the SYNTAM method with the NASA DeepBlue algorithm. The AOT values of more than 90% of pixels derived by both methods are very close. This clearly shows that AOT data from SYNTAM are very close to the AOT data set from the NASA DeepBlue algorithm in cloud-free areas. The synergic use of both the SYNTAM and DeepBlue algorithms could produce AOT values over much greater land areas. © 2012 Copyright Taylor and Francis Group, LLC. Source


Liu T.,Chinese Academy of Sciences | Wu X.,Guangdong Electronics Industry Institute | Lian Y.,Xinyutong Technology Co.
Proceedings of 2015 4th International Conference on Computer Science and Network Technology, ICCSNT 2015 | Year: 2015

Kitchen safety in catering industry has a direct impact on people's health and life safety. In this paper, a kitchen video monitoring system named sunshine kitchen is designed and constructed, which is based on cloud computing and video data analysis and provides services for the public, government, operators and other groups of people. One cloud data center is constructed as the supporting and shared modular for the application platforms, providing massive data collecting, storage and processing capacity. Three business platforms, public service platform, video monitor platform and operation platform, are established for public usage, so that the key issues of food safety such as illegal transaction and behavior, the problem of food resources, can be identified and exposed to public. Test results show that the platform can handle hundreds of user concurrent access. We hope this case will be promoted in more places. © 2015 IEEE. Source


Peng Z.,Huaqiao University | Wang T.,Huaqiao University | Alam Bhuiyan M.Z.,Temple University | Wu X.,Guangdong Electronics Industry Institute | Wang G.,Central South University
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2015

Recent years have witnessed the development of Camera Sensor Networks (CSNs), but most of existing researches are based on homogeneous CSNs, which are expensive for the transition from traditional Wireless Sensor Networks (WSNs). In this paper, we focus on the target tracking problem in Heterogeneous Mobile Camera Sensor Networks (HMCSNs) which consists of a large number of static common sensors and a small number of mobile camera sensors. The objective is to track the target dependably with maximized effective monitoring time and short moving distance. A novel tracking algorithm is proposed where multiple camera sensors cooperate with each other to track the target in a cascading scheme. It improves the effective monitoring time and shortens the moving distance of cameras. Moreover, the dependability is also improved even if the prediction about the target is not accurate. The effectiveness of the proposed algorithm is validated by simulation results, which show high monitoring ratio and comparatively short moving distance. © Springer International Publishing Switzerland 2015. Source


Liu T.,Chinese Academy of Sciences | Yue Q.,Guangdong Electronics Industry Institute | Ji T.,Chinese Academy of Sciences
WIT Transactions on Information and Communication Technologies | Year: 2014

Recently, more and more education activities and resources are concentrated on computers and networks. However, the vertical construction of the education systems has led to a lot of duplication of hardware resources and scattered data resources. Cloud computing has good prospects in the field of education, because it dramatically reduces the construction cost, brings data integration capability, and has efficient and flexible working models. In this paper, a case study of Dongguan education cloud computing platform is introduced. By describing the design and implementation of this platform, we hope this efficient, low-carbon, and low-cost construction model will be promoted in more education fields. © 2014 WIT Press. Source


Liu T.,Chinese Academy of Sciences | Yue Q.,Guangdong Electronics Industry Institute | Wu X.,Guangdong Electronics Industry Institute
Proceedings of 2015 IEEE International Conference on Computer and Communications, ICCC 2015 | Year: 2015

Based on the actual needs of Smart Humen Port, a logistics cloud platform for Dongguan Port of Humen is constructed. Using cloud computing techniques, the platform implements the resource integration of physical servers, storage devices and a variety of heterogeneous applications, and provides enterprises public infrastructure services, basic business systems and supporting systems, and enterprise portal online customizing and building services. This paper describes the software architecture and design ideas of the plat-form, and gives the integrated application cases. Experimental results show that the constructed platform has good availability and load balancing property. We hope this case study can help others to complete similar systems and this efficient, low-carbon, and low-cost construction model will be promoted in more fields. © 2015 IEEE. Source

Discover hidden collaborations