基于MBLBPV算法的布匹瑕疵检测方法

针对布匹瑕疵检测,在传统局部二值模式(Local Binary Pattern, LBP)与局部二值模式方差(LBP Variance,LBPV)的基础上,提出一种基于多尺度分块局部二值模式方差(Multi-Scale Block Local Binary Patterns Variance, MBLBPV)的检测算法。首先,采用适当尺度大小的子区域灰度均值代替单像素灰度值,提取LB P特征,以降低噪声影响;然后,融合图像区域对比度信息,并将其作为编码值的权重,提取图像MBLBPV特征,并基于该特征实现瑕疵的检测。实验结果表明,相对于传统方法,MBLBPV抗噪力强、检测正确率更高。     

DP-Share: Privacy-Preserving Software Defect Prediction Model Sharing Through Differential Privacy

In current software defect prediction (SDP) research, most previous empirical studies only use datasets provided by PROMISE repository and this may cause a threat to the external validity of previous empirical results. Instead of SDP dataset sharing, SDP model sharing is a potential solution to alleviate this problem and can encourage researchers in the research community and practitioners in the industrial community to share more models. However, directly sharing models may result in privacy disclosure, such as model inversion attack. To the best of our knowledge, we are the first to apply differential privacy (DP) to privacy-preserving SDP model sharing and then propose a novel method DP-Share, since DP mechanisms can prevent this attack when the privacy budget is carefully selected. In particular, DP-Share first performs data preprocessing for the dataset, such as over-sampling for minority instances (i.e., defective modules) and conducting discretization for continuous features to optimize privacy budget allocation. Then, it uses a novel sampling strategy to create a set of training sets. Finally it constructs decision trees based on these training sets and these decision trees can form a random forest (i.e., model). The last phase of DP-Share uses Laplace and exponential mechanisms to satisfy the requirements of DP. In our empirical studies, we choose nine experimental subjects from real software projects. Then, we use AUC (area under ROC curve) as the performance measure and holdout as our model validation technique. After privacy and utility analysis, we find that DP-Share can achieve better performance than a baseline method DF-Enhance in most cases when using the same privacy budget. Moreover, we also provide guidelines to effectively use our proposed method. Our work attempts to fill the research gap in terms of differential privacy for SDP, which can encourage researchers and practitioners to share more SDP models and then effectively advance the state of the art of SDP.     

4-Mask工艺Cu腐蚀分析及改善研究

针对4-Mask工艺铜数据线腐蚀造成的锯齿状不良现象进行系统研究,发现铜腐蚀发生的工艺步骤和机理,并找到有效的措施。首先,通过显微镜对每道刻蚀工艺后铜数据线形貌进行观测,确定铜腐蚀发生的工艺步骤。接着,通过扫描电子显微镜和X射线电子能谱测量腐蚀生产物成分,对腐蚀机理提出合理解释。最后在铜腐蚀发生机理基础上,提出有效的改善措施。铜腐蚀是在有源半导体层干刻和光刻胶的灰化综合作用下发生,其主要产物为氧化铜、氯化铜。通过先进行灰化工艺然后进行有源半导体层干刻的工艺措施,在铜数据线两侧形成氧化铜保护膜,可以彻底改善铜腐蚀。改善措施可以解决铜腐蚀的问题,彻底消除铜数据线锯齿状不良。     

Defect Engineering in Semiconductors: Manipulating Nonstoichiometric Defects and Understanding Their Impact in Oxynitrides for Solar Energy Conversion

Nonstoichiometric defects, as manifested by slight deviation of elemental compositions from chemical formulas, are common yet highly important in solid materials. Oxynitrides with a relatively large O/N ratio variation are theoretically predicted to change their electronic structures and charge transport behaviors with these nonstoichiometric defects. However, little experimental effort is devoted to understanding the impact of such nonstoichiometric defects regarding varied O/N ratios in oxynitrides for solar water splitting. The main reason is the lack of suitable oxynitride research models for aforementioned nonstoichiometric defect study without interference from other factors. Using TaON as a prototypical material, finely tuned O/N ratios can dramatically influence its photoresponse. In‐depth analysis further reveals a significant impact of nonstoichiometric defects (O/N ratios) on TaON's charge carrier densities, charge separation, and transport. Finally, manipulating nonstoichiometric defects of O/N ratios demonstrates its ability to control the space charge layer width and film conductivity of TaON photoanodes for high efficiency water splitting. Therefore, a fine understanding and control of nonstoichiometric defects would be highly important for future development of high efficiency oxynitrides for water splitting.     

基于改进Gabor优化选择的布匹瑕疵检测方法

针对传统Gabor优化选择方法用于布匹瑕疵检测时准确率低、鲁棒性差的缺点,提出了改进的优化选择方法,通过瑕疵图像与标准图像Gabor滤波后分块子图均值差平方和的代价函数实现优化选择。设置一组不同方向和尺度的Gabor滤波器并提取标准图像滤波后相关参数,通过改进的优化选择方法实现滤波后瑕疵图像的最优选择,利用迭代式阈值分割对最优滤波后图像进行二值分割,根据分割后图像的像素信息检测是否含有瑕疵并输出瑕疵信息。实验验证该方法,并与传统优化选择方法对比分析,结果表明该方法运算量较少,且检查性能高,可满足在线检测要求。     

输油气管道非开挖弱磁检测

针对长输油气管道随着服役时间增长出现腐蚀泄露甚至安全问题,提出埋地管道非开挖弱磁检测技术,能实现埋地管道在役非开挖检测。简述了弱磁检测技术和磁梯度法的基本原理以及基于统计学原理的缺陷判定方法,并对埋地管道作了失效分析。利用弱磁检测仪器对有缺陷的埋地管道作了试验,验证了该检测技术的可行性。结果表明,该检测方法对埋地管道腐蚀检测有很好的检测效果和检出率,能实现埋地管道在役时的非开挖检测以及腐蚀缺陷的二维成像。     

开关柜局部放电信号降噪与缺陷识别研究

文中采用暂态地电压法(TEV)进行检测,设计了四种典型的缺陷模型并搭建试验平台,分别对局部放电缺陷模型进行了实验。由于变电站现场环境复杂,需要对采集的信号进行信号降噪。针对以往小波降噪都是按照经验采取固定的分解层数的问题,提出一种Mallat算法结合最优分解层数自适应算法对含噪信号进行分离与重构,结果显示该算法可以很好地滤除噪声。对重构后的局放信号提取八种时域特征参数,并采用BP神经网络对开关柜局部放电的类型进行识别,当误差准确率δ=0.002时,放电类型的识别正确率最高,能够达到97%。     

Electronic Structure and Optoelectronic Properties of Bismuth Oxyiodide Robust against Percent‐Level Iodine‐, Oxygen‐, and Bismuth‐Related Surface Defects

In the search for nontoxic alternatives to lead‐halide perovskites, bismuth oxyiodide (BiOI) has emerged as a promising contender. BiOI is air‐stable for over three months, demonstrates promising early‐stage photovoltaic performance and, importantly, is predicted from calculations to tolerate vacancy and antisite defects. Here, whether BiOI tolerates point defects is experimentally investigated. BiOI thin films are annealed at a low temperature of 100 °C under vacuum (25 Pa absolute pressure). There is a relative reduction in the surface atomic fraction of iodine by over 40%, reduction in the surface bismuth fraction by over 5%, and an increase in the surface oxygen fraction by over 45%. Unexpectedly, the Bi 4f_7/2 core level position, Fermi level position, and valence band density of states of BiOI are not significantly changed. Further, the charge‐carrier lifetime, photoluminescence intensity, and the performance of the vacuum‐annealed BiOI films in solar cells remain unchanged. The results show BiOI to be electronically and optoelectronically robust to percent‐level changes in surface composition. However, from photoinduced current transient spectroscopy measurements, it is found that the as‐grown BiOI films have deep traps located ≈0.3 and 0.6 eV from the band edge. These traps limit the charge‐carrier lifetimes of BiOI, and future improvements in the performance of BiOI photovoltaics will need to focus on identifying their origin. Nevertheless, these deep traps are three to four orders of magnitude less concentrated than the surface point defects induced through vacuum annealing. The charge‐carrier lifetimes of the BiOI films are also orders of magnitude longer than if these surface defects were recombination active. This work therefore shows BiOI to be robust against processing conditions that lead to percent‐level iodine‐, bismuth‐, and oxygen‐related surface defects. This will simplify and reduce the cost of fabricating BiOI‐based electronic devices, and stands in contrast to the defect‐sensitivity of traditional covalent semiconductors.