A novel image tamper localization and recovery algorithm based on watermarking technology

In this paper, an effective self-embedding fragile watermarking scheme is presented. The watermark is generated by encoding the DCT coefficients of each 2 × 2 block and embedded into another block according to the block mapping. A non-linear chaotic sequence is used for generating the block mapping which can enhance the security of the algorithm. An improved tamper localization and recovery algorithm are performed. The experiment results show that the watermarked image has a high average peak signal-to-noise ratio. Furthermore, the tamper region can be successfully localized and exactly recovered, even if for the content-only tampering.     

The recognition and localization of insulators adopting SURF and IFS based on correlation coefficient

The recognition and localization of insulators is a crucial process of insulators’ failure detection. Aiming to identify insulators in test aerial image under the complicated background with no template and segmentation, a novel method of the recognition and localization of insulators based on SURF (Speeded Up Robust Features) and IFS (Intuitionistic Fuzzy Set) based on correlation coefficient is proposed as following: SURF is used to extract a set of feature points of the test aerial image. Then use IFS based on correlation coefficient to partition points into k classes. Find all the connected regions for each class, and calculate minimum circumscribed rectangle and shape characteristic value for each region to identify insulators, and localize them by rectangular boxes. Thus we can detect all the insulators from test aerial image correctly. Experimental results show that the proposed approach can obtain high precision in the recognition and localization of insulators.     

Influence of structural defects on the magnetization process in high-magnetic fields in the Fe61Co10Y8Nb1B20 alloy in the form of ribbons and plates

Amorphous samples of the Fe₆₁Co₁₀Y₈Nb₁B₂₀ alloy in the form of a ribbon with thickness of approximately 30 μm and in the form of a plate with thickness of 0.5 mm have been investigated. Plates were obtained by a suction-casting method, and ribbons by a single roller melt-spinning method. Microstructural studies, using X-ray spectroscopy, confirmed that samples in the as-quenched state were fully amorphous. Measurements carried out using a vibrating sample magnetometer allowed values of coercivity and saturation magnetization to be obtained; it also facilitated the estimation of size and nature of structural defects in the investigated samples. All of the investigated samples of this alloy in the as-quenched state have high values of saturation magnetization of greater than 1.25 T.     

One-dimensional Anderson model with dichotomic correlation in the presence of external electric field

A one-dimensional diagonal tight binding electronic system with dichotomic correlated disorder in the presence of external d.c field is investigated. It is found numerically that the conductance distribution obeys fairly well to log-normal distribution in weak disorder strength in localized regime, which indicates validity of single parameter scaling theory in this limit. Contrary to the universal cumulant relation C ₁ = 2C ₂ in the absence of d.c. field, we demonstrated numerically that C ₁ ≫ 2C ₂ in the presence of the field in localized regime. We interpret this result as suppression of the fluctuation effects by the external field. In addition, it is obtained that the quantity NF _c , here N is the system size and F _c is the crossover field, decreases as the as the system energy E increases. Moreover, we find numerically a simple linear relation between the average logarithm of the conductance 〈ln(g)〉 and the field strength as 〈ln(g)〉 = C(N, λ)F, here C(N, λ) is a constant for particular values of N and λ, which is the Poisson parameter of the dichotomic process.     

无线传感器网络节点定位技术综述

无线传感器网络在许多领域有着重要的科研和使用价值,网络中传感器节点自身定位可为无线传感器网络的很多应用提供基础信息,是重要研究方向之一.从无线传感器网络节点定位技术的研究意义与应用价值出发,介绍了节点定位技术的基本原理与方法,并讨论了定位算法的评价标准,最后对节点定位技术的发展方向进行了展望.     

基于目标检测网络的轮对踏面缺陷检测方法

针对传统图像处理算法难以快速、准确识别轮对踏面缺陷的问题,提出一种采用双深度神经网络对轮对踏面缺陷进行检测的算法。该双网络分为踏面提取网络与缺陷识别网络。根据踏面为大目标的特点,分析与测试SSD网络,并用该网络提取轮对图像中的踏面区域。为提高踏面缺陷识别效率,在提取出踏面图像后,针对踏面缺陷属于中、小目标的特点,对YOLOv3网络结构进行优化得到M-YOLOv3。实验测试表明:提取踏面区域时,SSD算法的精度均值(AP)最高,达99.8%;识别踏面缺陷时,M-YOLOv3的AP达89.9%,相较于原始YOLOv3,单张图像计算耗时减少7.1%,同时AP仅有0.6%的损耗。结果表明,所提算法具有较高的检测准确率。     

Contamination due to process gases

A survey is given of the BOC Group technical research programs, whose purpose is to relate purity in gas processing to ULSI device parameters. Results from the following research programs are presented: how inert gas purity affects the Ti silicide for IGFET metallization; the effects of argon versus nitrogen used in silicon gate oxidation process steps; and preliminary work on a new gas analytical tool, Atmospheric Pressure Ionization Mass Spectrometry (APIMS), which allows measurement of sub ppb impurity levels in processing gases.     

PCB企业如何进行过程控制

本文通过探讨六大因素对PCB制造各工序的影响程度,简要阐述了在PCB制造过程中如何进行质量控制。