A simple co-precipitation approach taking place between Ln3＋, Sr2＋ cations and F anions, led to the formation of nanocrystalline Eu3＋ doped Sr2LnF7 （Ln-La and Gd） complex fluorides. The reaction was carried out i...A simple co-precipitation approach taking place between Ln3＋, Sr2＋ cations and F anions, led to the formation of nanocrystalline Eu3＋ doped Sr2LnF7 （Ln-La and Gd） complex fluorides. The reaction was carried out in the presence of polyeth- ylene glycol, PEG 6000 as a surfactant/surface modifier, providing small size and homogeneity of the products. The synthesized compounds were composed of small nanoparticles with an average size of 15 nm. All obtained Eu3＋ doped compounds exhibited an intensive red luminescence. In the case of gadolinium based compounds, the energy transfer phenomena could be observed from Gd3＋ ions to Eu3＋ ions. In order to study the structure and morphology of the synthesized fluorides, powder X-ray diffraction （XRD） and transmission electron microscopy （TEM） measurements were performed. Also FT-IR spectra of the products were re- corded, revealing the presence of PEG molecules on the nanoparticles suN＇ace. A spectrofluorometry technique was applied to examine optical properties of the synthesized nanoparticles. Excitation and emission spectra as well as luminescence decay curves were measured and analysed. The performed analysis revealed a red luminescence, typical for the Eu3＋ ion situated in the inorganic, highly symmetric matrix. Concentration quenching phenomena and lifetimes shortening, together with an increasing of the Eu3＋ doping level, were observed and discussed. Judd-Ofelt analysis was also performed for all doped samples, in order to support the registered spectroscopic data and examine in details structural and optoelectronic properties of the synthesized nanomaterials.展开更多
针对人脸特征点定位的精确性对人脸识别系统精确性的影响,本文在受约束的局部模型(constrained local models,CLM)基础上,主要研究人脸特征点定位算法——受约束的局部神经域模型(constrained local neural fields,CLNF)算法。考虑每个p...针对人脸特征点定位的精确性对人脸识别系统精确性的影响,本文在受约束的局部模型(constrained local models,CLM)基础上,主要研究人脸特征点定位算法——受约束的局部神经域模型(constrained local neural fields,CLNF)算法。考虑每个patch模型(特征点检测器)的可靠性,CLNF结合局部神经域(local neural field,LNF)的patch模型,在拟合过程中,由原来的正则化特征点均值偏移(regularised landmark mean shift,RLMS)改为采用不均匀的正则化特征点均值偏移方法进行人脸拟合,同时,在人脸数据集Multi-PIE上进行实验,并对比分析两种模型。分析结果表明,CLNF定位算法在平均用时、成功率及误差率方面都具有明显优势,证明CLNF的LNF patch模型在人脸特征点拟合的精确性相对于CLM有明显提高。该技术拟合速度更快,拟合准确率更高,能够使人脸识别技术更加精确,具有更大的优势。该研究具有广泛的应用前景。展开更多
基金Project supported by Polish Ministry of Science and Higher Education(Diamond Grant"Nr DI2011 011441)
文摘A simple co-precipitation approach taking place between Ln3＋, Sr2＋ cations and F anions, led to the formation of nanocrystalline Eu3＋ doped Sr2LnF7 （Ln-La and Gd） complex fluorides. The reaction was carried out in the presence of polyeth- ylene glycol, PEG 6000 as a surfactant/surface modifier, providing small size and homogeneity of the products. The synthesized compounds were composed of small nanoparticles with an average size of 15 nm. All obtained Eu3＋ doped compounds exhibited an intensive red luminescence. In the case of gadolinium based compounds, the energy transfer phenomena could be observed from Gd3＋ ions to Eu3＋ ions. In order to study the structure and morphology of the synthesized fluorides, powder X-ray diffraction （XRD） and transmission electron microscopy （TEM） measurements were performed. Also FT-IR spectra of the products were re- corded, revealing the presence of PEG molecules on the nanoparticles suN＇ace. A spectrofluorometry technique was applied to examine optical properties of the synthesized nanoparticles. Excitation and emission spectra as well as luminescence decay curves were measured and analysed. The performed analysis revealed a red luminescence, typical for the Eu3＋ ion situated in the inorganic, highly symmetric matrix. Concentration quenching phenomena and lifetimes shortening, together with an increasing of the Eu3＋ doping level, were observed and discussed. Judd-Ofelt analysis was also performed for all doped samples, in order to support the registered spectroscopic data and examine in details structural and optoelectronic properties of the synthesized nanomaterials.
文摘针对人脸特征点定位的精确性对人脸识别系统精确性的影响,本文在受约束的局部模型(constrained local models,CLM)基础上,主要研究人脸特征点定位算法——受约束的局部神经域模型(constrained local neural fields,CLNF)算法。考虑每个patch模型(特征点检测器)的可靠性,CLNF结合局部神经域(local neural field,LNF)的patch模型,在拟合过程中,由原来的正则化特征点均值偏移(regularised landmark mean shift,RLMS)改为采用不均匀的正则化特征点均值偏移方法进行人脸拟合,同时,在人脸数据集Multi-PIE上进行实验,并对比分析两种模型。分析结果表明,CLNF定位算法在平均用时、成功率及误差率方面都具有明显优势,证明CLNF的LNF patch模型在人脸特征点拟合的精确性相对于CLM有明显提高。该技术拟合速度更快,拟合准确率更高,能够使人脸识别技术更加精确,具有更大的优势。该研究具有广泛的应用前景。
