A filter bank method to construct rotationally invariant moments for pattern recognition

We propose multiresolution filter bank techniques to construct rotationally invariant moments. The multiresolution pyramid motivates a simple but efficient feature selection procedure based on a combination of a pruning algorithm, a new version of the Apriori mining techniques and the partially supervised fuzzy C-mean clustering. The recognition accuracy of the proposed techniques has been tested with the reference to conventional methods. The numerical experiments, with more than 50,000 images taken from standard image datasets, demonstrate an accuracy increase ranging from 5% to 27% depending on the noise level.     

Multiplicity of steady states in FCC units: effect of operating conditions

The FCC dynamics has been studied to determine the effect of operating conditions on multiplicity of steady states. The proposed model is capable to analyze several dynamic features to characterize the FCC process. Three combustion modes are analyzed: Partial combustion, total combustion and transition mode. At least three steady states have been determined in both partial and complete combustion modes. Furthermore, it is established in this paper that operating conditions are incapable to change the number of steady states. It is believed that a similar analysis should be applied to identify the effect on the multiplicity of physical parameters such as oxidation rates.     

基于颜色空间映射的自适应足球机器人视觉系统

为了解决足球机器人比赛中视觉信息检测问题,给出基于嵌入式多处理器体系结构的视觉系统框架,并研究了一种自适应阈值目标分割方法。该方法将视觉系统采集的图像从YCbCr颜色空间转换为一维颜色映射图像,并提出基于竞争学习的空间映射方法。同时使用小波变换方法来平滑颜色映射空间直方图和降低阈值搜索维数。然后使用基于映射图像直方图动态搜索阈值的方法来分割图像。最后,通过实验表明研究的视觉系统的有效性,能够实时准确地完成视觉处理任务。     

Analysis on the spatial ecological distribution model of landscape plant community

Because the spatial pattern of plant communities in garden landscapes usually exhibits highly non-random distribution characteristics, the accuracy of the analysis results is not high. For this purpose, a spatial ecological distribution model of landscape landform plant communities is designed. Based on the remote sensing images of landscape plant communities, the study area was obtained. The grassland communities in this area were classified and calculated by geostatistics, and the semi variogram values of typical samples of different vegetation types were obtained. According to the calculation results, the spatial terrain factors of landscape plant community are extracted, and the annual NDVI value is taken as the ecological vitality index of landscape green vegetation, the NDVI level is divided, the spatial ecological distribution model of landscape plant community is constructed, and the evaluation index system is generated, so as to complete the spatial ecological distribution model analysis of landscape plant community. The results show that the method of this study accurately analyzes the area of coniferous and broad-leaved mixed forest, mountain meadow, road, construction land, water area land and slope grade, which is slightly different from the actual value, indicating that the method has certain reliability.     

Paradigm and performance analysis of distributed frequent itemset mining algorithms based on Mapreduce

FrequentItemsetMining (FIM) is one of the most important data mining tasks and is the foundation of many data mining tasks. In Big Data era, centralized FIM algorithms cannot meet the needs of FIM for big data in terms of time and space, so Distributed Frequent Itemset Mining (DFIM) algorithms have been designed to meet the above challenges. In this paper, LocalGlobal and RedistributionMining which are two main paradigms of DFIM algorithm are discussed; Two algorithms of these paradigms on MapReduce named LG and RM are proposed while MapReduce is a popular distributed computing model, and also the related work is discussed. The experimental results show that the RM algorithm has better performance in terms of computation and scalability of sites, and can be used as the basis for designing the DFIM algorithm based on MapReduce. This paper also discusses the main ideas of improving the DFIM algorithms based on MapReduce.     

Phase equilibria of the Al2O3–CaO–SiO2-(0%, 5%, 10%) MgO slag system for non-metallic inclusions control

Al₂O₃–CaO-(MgO–SiO₂) inclusions are one of the dominant inclusions in Al-deoxidized spring steel, the compositions changes of which are closely related to refining slags and deoxidization process. The Al₂O₃–CaO–SiO₂–MgO system can represent the primary ingredients of the Al₂O₃–CaO inclusions. According to analyzed compositions and predicted liquidus temperature ranges of inclusions and refining slag, equilibra experiments under high temperature, water quenching technique and subsquent electron probe X-ray microanalysis (EPMA) has been conducted to ascertain detailed thermodynamic database for inclusions control. Liquidus temperatures within the dominant phase fields of Ca₃SiO₅, Ca₂SiO₄, CaAl₂O₄, Ca₃Al₄O₉, spinel and MgO with the intervals of 20 °C from 1350 to 1560°C were identified. To further promote inclusions control, the influences of mass ratios of Mass(Al2O3)Mass(Al2O3+SiO2+CaO) and MgO contents on equilibrated phases and liquidus temperature changes have been explored. To further enhance modification levels of Al₂O₃–CaO-(MgO–SiO₂) system inclusions, it is suggested that refining time could be suitably prolonged.     

多频段脑功能网络融合的阿尔茨海默病分类

将图论与机器学习方法相结合的阿尔茨海默病计算机辅助中,脑网络的构建大多是基于滤波去噪后的全频段BOLD信号匹配,忽略了不同脑活动信息的差异.因此,本文提出了一种多频段脑功能网络融合模型.首先将离散小波变换应用于BOLD信号中,得到不同频域下的体素信号,而后计算同频信号的相关性,获取不同频段下相关矩阵.而后计算所有矩阵的网络特征,在特征选择后基于SVM对患者进行分类.从实验结果可以看出,分频下的脑功能网络特征与未分频网络相比能在一定程度上提高分类的准确性;体素级网络由于可以更加详细的表达脑网络的变化,其分类效果要优于脑区级.