Enhanced photoresponse performance in Ga/Ga_2O_3 nanocomposite solar-blind ultraviolet photodetectors

In the present work, we explore the solar-blind ultraviolet(UV) photodetectors(PDs) with enhanced photoresponse,fabricated on Ga/Ga_2O_3 nanocomposite films. Through pre-burying metal Ga layers and thermally post-annealing the laminated Ga2 O3/Ga/Ga_2O_3 structures, Ga/Ga_2O_3 nanocomposite films incorporated with Ga nanospheres are obtained. For the prototype PD, it is found that the photocurrent and photoresponsivity will first increase and then decrease monotonically with the thickness of the pre-buried Ga layer increasing. Each of all PDs shows a spectrum response peak at 260 nm, demonstrating the ability to detect solar-blind UV light. Adjustable photoresponse enhancement factors are achieved by means of the surface plasmon in the nanocomposite films. The PD with a 20 nm thick Ga interlayer exhibits the best solar-blind UV photoresponse characteristics with an extremely low dark current of 8.52 p A at 10-V bias, a very high light-to-dark ratio of ～ 8 × 10~5, a large photoresponsivity of 2.85 A/W at 15-V bias, and a maximum enhancement factor of ～ 220. Our research provides a simple and practical route to high performance solar-blind UV PDs and potential applications in the field of optoelectronics.     

光谱图像技术结合SAM算法识别自然场景下的成熟柑橘

为了实现采摘机器人在复杂的自然场景下正确识别树上果实,来完成果实采摘,研究了不同环境下柑橘的识别方法.针对复杂的自然环境的影响及传统方法的局限性,在可见光和近红外区域择选5个特征波长滤波片,采集得到5幅滤波后的图像,并利用光谱角分类算法完成柑橘识别.试验结果表明,在光照角度、光照强度等不同条件下,柑橘的识别准确度达到96%.研究表明,滤波片光谱图像技术结合光谱角分类算法可以有效地识别自然场景下的成熟柑橘.     

利用高光谱图像技术评判茶叶的质量等级

针对茶叶品质无损检测时内外品质难以同时兼顾的问题,利用高光谱图像技术检测茶叶质量.设计一套基于光谱仪的高光谱图像系统采集数据;通过主成分分析,从海量数据中优选出三个波长段的特征图像;从每个特征图像中分别提取平均灰度级、标准方差、平滑度、三阶矩、一致性和熵等6个基于统计矩的纹理特征参量,每个样本共有18个特征变量;再通过主成分分析对这18个特征变量进行压缩,提取8个主成分因子建立基于反向传播神经网络的茶叶等级判别模型.模型训练时的总体回判识别率为97%;预测时总体识别率为94%.结果表明,高光谱图像技术可以用于茶叶质量等级水平的评判.     

近红外光谱法快速检测绿茶中儿茶素的含量

提出了一种应用傅里叶近红外漫反射光谱分析技术快速检测茶叶中主要儿茶素含量的新方法.首先获取茶叶在10000～4000 cm<'-1>范围的近红外漫反射光谱,然后以高效液相色谱分析值作参考值.采用偏最小二乘法建立茶叶中表没食子儿茶素没食子酸酯(EGCG)、表儿茶素没食子酸酯(ECG)和表没食子儿茶素(EGC)含量的定量分析模型.通过交互验证方法来优化模型的主成分数和所采用的光谱预处理方法.EGCG、ECG和EGC三个模型预测值和参考值问的相关系数分别为0.9800、0.9763和0.9853,预测均方根误差分别为0.3509、0.1147和0.1365.研究结果表明,近红外光谱技术可成功地检测茶叶中EGCG,ECG和EGC的含量.