基于改进的背景模型的图像识别算法研究

当图像亮度不均匀、对比度低时,提取图像前景较困难。为此,提出一种图像分割方法,结合正弦基函数和绝对值距离测度构建背景模型,依据优化理论和迭代法求解背景模型,通过比较背景模型中各像素点亮度与实际图像中各像素点亮度来判别各像素点是背景还是前景。为应对图像亮度不均匀的情况,在图像分割前对图像进行分块,在分块图像中依据背景模型或相邻分块背景相似度进行图像分割。实验结果表明,在普适性方面,相对于经典的模糊C均值法和OTSU法,该方法的分割误差小,尤其是对亮度不均匀和对比度低的图像;在掌纹图像分割应用方面,与迭代线跟踪法和模糊粗糙集法相比,该方法的错误率低、信噪比高、处理时间短。最后将提出的分割算法应用在人脸识别上,实验结果表明了该算法的先进性。     

Supervisory control based on minimal cuts and Petri net sub-controllers coordination

This paper addresses the synthesis of Petri net (PN) controller for the forbidden state transition problem with a new utilisation of the theory of regions. Moreover, as any method of control synthesis based on a reachability graph, the theory of regions suffers from the combinatorial explosion problem. The proposed work minimises the number of equations in the linear system of theory of regions and therefore one can reduce the computation time. In this paper, two different approaches are proposed to select minimal cuts in the reachability graph in order to synthesise a PN controller. Thanks to a switch from one cut to another, one can activate and deactivate the corresponding?PNcontroller. An application is implemented in a flexible manufacturing system to illustrate the present method. Finally, comparison with previous works with experimental results in obtaining a maximally permissive controller is presented.     

Fuzzy distributed cooperative tracking for a swarm of unmanned aerial vehicles with heterogeneous goals

This article proposes a systematic analysis for a tracking problem which ensures cooperation amongst a swarm of unmanned aerial vehicles (UAVs), modelled as nonlinear systems with linear and angular velocity constraints, in order to achieve different goals. A distributed Takagi–Sugeno (TS) framework design is adopted for the representation of the nonlinear model of the dynamics of the UAVs. The distributed control law which is introduced is composed of both node and network level information. Firstly, feedback gains are synthesised using a parallel distributed compensation (PDC) control law structure, for a collection of isolated UAVs; ignoring communications among the swarm. Then secondly, based on an alternation-like procedure, the resulting feedback gains are used to determine Lyapunov matrices which are utilised at network level to incorporate into the control law, the relative differences in the states of the vehicles, and to induce cooperative behaviour. Eventually stability is guaranteed for the entire swarm. The control synthesis is performed using tools from linear control theory: in particular the design criteria are posed as linear matrix inequalities (LMIs). An example based on a UAV tracking scenario is included to outline the efficacy of the approach.     

Guarding against online threats: why entrepreneurs take protective measures

Information technology is becoming increasingly important for entrepreneurs. Protecting their technical infrastructure and stored data is, therefore, also becoming more vital. Nevertheless, research into the safety and security of entrepreneurs in general, and online threats targeted at entrepreneurs in particular, is still limited. This study investigates to what extent self-employed entrepreneurs protect themselves against online threats, and what motivates them to do so. Based on secondary analyses on data collected from 1622 Dutch entrepreneurs, we observe that the majority implement technical and personal coping measures. By adopting protection motivation theory as a theoretical basis for our study, we found that coping appraisal explains the adoption of protective measures. Entrepreneurs are likely to implement protective measures when they believe a measure is effective, when they are capable of using internet technology, when their attitude towards online protection is positive and when they believe they are responsible for their own online security. Although the secondary analysis provides some limitations, this study offers new insights into the usage of protective measures and the motivation for taking them. These insights can help to improve information security campaigns.     

Multi‐Unit Bayesian Auction with Demand or Budget Constraints

We consider the problem of revenue maximization on multi‐unit auctions where items are distinguished by their relative values; any pair of items has the same ratio of values to all buyers. As is common in the study of revenue maximizing problems, we assume that buyers' valuations are drawn from public known distributions and they have additive valuations for multiple items. Our problem is well motivated by sponsored search auctions, which made money for Google and Yahoo! in practice. In this auction, each advertiser bids an amount b_i to compete for ad slots on a web page. The value of each ad slot corresponds to its click‐through‐rate, and each buyer has her own per‐click valuations, which is her private information. Obviously, a strategic bidder may bid an amount that is different with her true valuation to improve her utility. Our goal is to design truthful mechanisms avoiding this misreporting. We develop the optimal (with maximum revenue) truthful auction for a relaxed demand model (where each buyer i wants at most d_i items) and a sharp demand model (where buyer i wants exactly d_i items). We also find an auction that always guarantees at least half of the revenue of the optimal auction when the buyers are budget constrained. Moreover, all of the auctions we design can be computed efficiently, that is, in polynomial time.     

Continuance usage of corporate SNS pages: A communicative ecology perspective

This study develops a research model that explains the continuance of corporate social network service (SNS) pages. In order to find antecedents that influence end users’ continuance intention, this study adopts communicative ecology theory from media and communication studies. Results indicate that the quality of social interaction is the most salient feature, as indicated by the fact that it has the highest impact on perceived usefulness and enjoyment. Moreover, content quality also increases both perceived usefulness and enjoyment, while SNS platform quality and service quality increase perceived usefulness only. In addition, perceived network size moderates the impact of social interaction quality on perceived enjoyment.     

Judging online peer-to-peer lending behavior: A comparison of first-time and repeated borrowing requests

The past decade has witnessed a growing number of business models that facilitate economic exchanges between individuals with limited institutional mediation. One of the important innovative business models is online peer-to-peer lending, which has received wide attention from government, industry, and researchers. Using the signaling theory, we compare the effects of various signals on the likelihood of successful funding in three models (i.e., first-time borrowing, repeated borrowing without prior lending, and repeated borrowing with prior lending). Using data collected from PPDAI.com, we verify the three proposed models by employing logistic regression. Results and implications are analyzed and discussed.     

Representations of lesbian and bisexual women’s sexual and relational health in online video and text-based sources

The Internet is the preferred source of health information followed by healthcare providers, friends and family, print media, and television (Baxter & Egbert, 2008). Yet, previous research about online sexual health resources for LGBTQ (lesbian, gay, bisexual, transgendered, queer) groups, especially lesbian and bisexual (LB) women, primarily consists of introductions and reviews of related websites rather than critique of the quality of the information present (e.g. Clark, 2003; Finlon, 2002; McKay, 2011). Thus, the goals of this study were to examine the LB women’s sexual and relational health content in online sources and identify the degree to which the online sources considered LB women’s identity and standpoint when presenting sexual and relational health messages. The analysis includes a comparison of text-based and video-based online messages related to LB women’s sexual and relational health. We argue that there is an overall lack of quality in the online sexual and relational health messages for LB women, and that websites are not comprehensive enough in their coverage of LB women’s sexual and relational health.     

Wald's mighty maximin: a tutorial

Ever since its appearance on the scene in the 1940s, in the then nascent field of statistical decision theory, Wald's maximin paradigm has played a vital role in many disciplines in the treatment of nonprobabilistic uncertainty, both as a tool of thought and as a practical instrument. In fact, in some fields, such as modern robust optimization, Wald's paradigm dominates the scene. It is important to note that this paradigm's preeminence in many fields continues unabated, despite its obvious limitations and the criticism that had been and continues to be leveled at it. So, in this tutorial we examine the methodological aspects of this stalwart of decision theory from the viewpoint of robust decision‐making, paying special attention to its obvious and not so obvious limitations and to its relation to other maximin paradigms.