Informatics Publishing Limited

Jun Li ¹, Hongmei Zhang ¹, Yuanjiang Liao ¹

Affiliations
1 School of Electron and Information Engineering, Ningbo University of Technology, CN

Abstract

This paper proposes a new approach to annotate image. First, in order to precisely model training data, shape context features of each image is represented as a bag of visual words. Then, we specifically design a novel optimized graph-based semi-supervised learning for image annotation, in which we maximize the average weighed distance between the different semantic objects, and minimize the average weighed distance between the same semantic objects. Training data insufficiency and lack of generalization of learning method can be resolved through OGSSL with significantly improved image semantic annotation performance. This approach is compared with several other approaches. The experimental results show that this approach performs more effectively and accurately.

Keywords

Image Retrieval, Image Semantic Annotation, Bag of Words (BoW), Semi-Supervised Learning.

Full Text

     

Hongmei Zhang ¹, Bairen Yang ¹, Aihui Chen ¹, Huixing Liang ¹

Affiliations
1 School of Environmental Science and Engineering, Yancheng Institute of Technology, Jian Jun Road No. 211, Yancheng, Jiangsu, 224051, CN

Abstract

The introduction of invasive species, Spartina alternifora, in the tidal marshes of Yancheng wetland has been considered deleterious to habitat quality. However, the extent to which the replacement of Suaeda salsa by S. alternifora affects the soil environment is unknown. In this study, we examined the soil physicochemical characteristics and microbial communities of rhizosphere and non-rhizosphere in soils adjacent to Spartina and Suaeda marshes. The dynamics of microbial community were similar in both the types of plants, and the ischolar_mains were conducive to microbial growth. The total number of microorganisms and microbial activity showed a seasonal fluctuation, increasing from spring through summer and then declining gradually to the lowest in spring. The data provided herein also indicated that values of organic carbon, biomass carbon, size of microbial populations, and activity were higher in Suaeda than in Spartina marsh, which suggests that differences in vegetation cover significantly affect the soil environment and microbial community. Therefore, we suggest that the policy to introduce S. alternifora into this area should be reconsidered.

Keywords

Microbial Activity, Soil Microbial Diversity, Fluorescein Diacetate, Spartina alternifora, Suaeda salsa.

Full Text

   

Hongmei Zhang ¹, Jiwen Wu ¹, Gao Yang ², Zhuo Chen ¹

Affiliations
1 School of Earth and Environment, Anhui University of Science & Technology, Huainan 232001, CN
2 Environment and Surveying Engineering College, Suzhou University, Suzhou 234000, CN

Abstract

With the growth of global population and intense land use changes, the problem of ecological vulnerability has become prominent. Assessment of ecological vulnerability bears significance in protection and restoration of the ecological environment. Assessment results for a given area can reveal regional characteristic differences in system vulnerability and provide basis for the rational use and protection of the natural environment. To analyse the ecological vulnerability in Fuzhou district in China and its causes, systematic assessment and zoning of ecosystem vulnerability were proposed. Using the remote sensing and geographic information system technologies, ecological vulnerability was spatially, explicitly, and comprehensively assessed. Results showed that the most vulnerable areas of Fuzhou district include Pingtan Comprehensive Experimental Area, Fuqing City, Fuzhou City and Changle City. Counties of Yongtai, Minhou and Minqing feature good vegetation coverages and low ecological vulnerability. In Fuzhou City, three zones, namely, Taijiang, Gulou and Cangshan Zones, feature high or extremely high percentage vulnerability. The other two zones comprise better ecosystems. Analysis of correlation coefficients between vulnerability value and various indices showed distinct factors of ecological vulnerability for each region. The main factors affecting the ecosystems in Fuqing, Jin’an Zone, and Changle City include climate, vegetation and soil, respectively, whereas the factors that simultaneously affect the ecosystem in Cangshan Zone consist of soil, land use, topography and vegetation. The key challenge in improving ecological vulnerability of these areas is optimization and coordination of land use/coverage under natural conditions. The results can provide basis and reference for future research and relevant formulation, such as the mode of selection of resource utilization and environmental protection, and improvement of key factors.

Keywords

Vulnerability Assessment, Ecological Threats, AHP Method, Remote Sensing, GIS.

Full Text