探索利用可见激光无损提取粗糙纸张上尿渍足迹的可行性，定量比较足迹显现效果。利用3种可见激光，配合使用11种可见波段滤光镜，对4类20种常见类型纸张上的尿渍足迹样本进行可见激光荧光照相，观察足迹显现效果；创建Matlab程序，引入灰度荧光亮度标准差指标比较足迹显出程度。发现20个尿渍足迹样本显现效果最好时的波段组合几乎各不相同，但可利用445 nm蓝色激光作为激发光，配合使用570 nm接收滤光镜，对所有样本的尿渍足迹进行整体最大程度的显现，样本尿渍足迹发现率100%；除黑色打印纸、复杂彩色背景报纸上的尿渍足迹样本外，样本尿渍足迹有效显出率90%。表明可见激光荧光照相能够用来显现粗糙纸张上的尿渍足迹，灰度荧光亮度标准差可以用来衡量尿渍足迹显出程度。

针对当前输电线路无人机巡检中树障隐患检测计算量大、效率较低的问题，设计了基于激光点云数据的电力线重建和树障隐患快速检测分析的方法。该方法首先根据高程分布特征、点云密度特征和倾斜角度特征提取电力线点云，分离各条电力线并将其矢量化，然后提取植被点云生成三维凸包，最后通过计算植被凸包点与电力线矢量的距离提高了树障检测效率，快速检测出该线路内的树障隐患信息。结果表明：电力线提取结果完整准确，电力线三维重建精度和树障隐患检测效率较高。该方法对输电线路地形、线路走向和点云密度等因素鲁棒性较好，大幅提高了输电线路树障隐患检测的质量和效率，为大规模复杂环境下的输电线路树障隐患检测应用提供了参考。

基于机器视觉的粗糙度测量方法大多是根据粗糙度关联指标建立预测模型，或者利用深度学习网络建立无指标预测模型，而这两类方法均存在着不足。一方面，人工设计指标的计算过程复杂，不利于在线检测。另一方面，深度学习模型则严重依赖大数据，数据量不足难以训练出有效的模型。针对以上问题，本文提出一种基于图神经网络的铣削表面粗糙度测量方法。该方法在训练阶段获取了自主学习的能力，而后仅需要少量铣削样本就能够完成铣削工件的粗糙度测量。试验结果表明，本文方法在铣削工件的粗糙度测量上不仅能够自动提取特征，而且表现出了较高的精度和良好的光照环境鲁棒性。

Objective of the study: We used fluorescence imaging methods of apoptosis and necrosis in human renal carcinoma A498 tumor cells in vitro to reveal the indicated forms of cell death under the combined effect of flavonoid-containing extract of Gratiola officinalis and cytostatic (cyclophosphamide). Materials and methods: The dyes were propidium iodide and acridine orange, which were used in the "alive and dead" test. This test helped us to identify the total number of dead cells in the forms of necrosis and apoptosis and the number of cells in which apoptosis had started, it was characterized by the appearance of apoptotic bodies or nucleus pyknosis. Results: We found the most pronounced cytotoxic activity at the ratio of extract of Gratiola officinalis and cyclophosphamide concentrations of 1:1. The number of living cells decreased when exposed to the ratio of extract and cytostatic concentrations of 2:1. When the ratio of concentration of the extract relative to the cytostatic increased to 3:1, the cytostatic activity of the extract began to appear, the total number of tumor cells decreased. The number of cells with nucleus pyknosis and the number of cells with apoptosis signs significantly increased at a 3:1 ratio of extract and cytostatic concentrations, which confirms the presence of pro-apoptotic activity of the studied combination. This trend indicates the dependence of a certain form of cell death (apoptosis, necrosis) on the ratio of extract and cytostatic doses, and it also demonstrates the cytostatic and cytotoxic effects of this combination. Conclusion: Fluorescence methods of investigation in the "alive and dead" test allowed us to visualize the forms of cell death of human kidney carcinoma A498 by combined exposure to the flavonoid-containing extract of Gratiola officinalis and cytostatic (cyclophosphamide) 24 h after exposure. We found that the combination with a concentration ratio of the extract and cyclophosphamide of 3:1 has the greatest effectiveness due to stimulation of the cytostatic effect and cytotoxic effect.

本试验以酿酒酵母(Saccharomyces cerevisiae)为材料, 研究玫瑰花提取物(RE)对砷胁迫下酵母细胞生长和凋亡的影响。试验以抗坏血酸(Vc)溶液为标准抗氧化对照物, 1.0 mmol/L亚砷酸钠(As)为酵母胁迫浓度, 分别设置了酵母对照组、酵母+RE(1.5 g/L)培养组、酵母+Vc(0.1 g/L)培养组、酵母+As(1.0 mmol/L)培养组、酵母+As+RE培养组和酵母+As+Vc培养组。酵母细胞培养24 h后, 利用噻唑兰(MTT)细胞增殖法、平板点样法、荧光素二乙酸/碘化丙锭(FDA/PI)染色法、4', 6-二脒基-2-苯基吲哚(DAPI)细胞核染色法和Annexin V-FITC/PI细胞凋亡流式检测法测定了酵母细胞的生长增殖与凋亡。试验结果显示, RE有效地缓解了砷对酵母细胞的损伤, 极显著提高了砷胁迫下酵母细胞的生长率和存活率(P<0.01), 极显著抑制了砷胁迫下酵母细胞的凋亡(P<0.01)。结果证实, RE对砷胁迫酵母细胞的凋亡有显著抑制作用, 在50%的1,1-二苯基-2-三硝基苯肼(DPPH)清除率下, 其对砷胁迫下酵母细胞的凋亡抑制率显著高于Vc(P<0.05)。因此, RE作为天然植物提取物, 因其绿色、无毒的作用机制, 将在对抗重金属污染、提高机体抗氧化力、缓解氧化损伤等方面有着更广阔的应用前景。

针对图片相似性比较的问题, 提出了一种基于数据场聚类的图片相似性比较方法。该方法首先把图片信息从RGB颜色模型转换到HSV颜色模型, 再把每个像素点投影到HSV三维空间, 通过数据场聚类算法提取主颜色, 最后依据两张图片的主颜色进行图片相似程度的判断。通过试验验证, 该算法具有较好相似性判断的效果。