Iris localization is to detect outer-and-inner boundaries of iris in an iris image. In the paper, an improved algorithm was proposed to quickly and effectively locate outer-and-inner boundaries. As for this algorithm, the first is to block an iris image and extract its sub-image blocks which cover pupil; the second is to set a binary threshold of pupil by adopting the method of Maximum Variance between Clusters; the third is to get the value outer-boundary-points of iris, on the basis of gray gradient of key Regions-of-interest; the last is to select some characteristic pixels in regions of interest respectively and fit outer-and-inner boundaries of iris according to curve fitting.

With the rapid development of domestic economy and the construction, the application of 10 kV power system is more and more extensive and the environment of the application is also more and more complicated. Aiming at this situation, in this paper, a new-type circuit breaker is proposed for the outer environment, which is based on the Water pressure up case configuration. The configuration is fully-enclosed, small size, light-weight, and it is applicable to high temperature and humid areas. In order to study characteristics of the permanent magnetic force (PMA), the paper adopts the electromagnetic analysis software ANSOFT to analyze its static and dynamic characteristics. The experiment results show that the performance of PMA is reliable. In order to analyze the insulting property of its enclosure design with new-style and small size, ANSOFT electric field is adopted to conduct simulation analysis and the experiment results show that the design and improvement measure are stable and reliable.

In current complex network environment, various kinds of attacks are mixed together forming the attack group. The diversity of the attack group leads to diverse attack signatures, which cannot be constrained by uniform conditions. Based on mixed attack signatures estimation model, the paper proposes a detection method for mixed and diverse attack groups. The paper classifies the different attacks in the mixed and diverse attack groups by use of attack constraint classification methods, builds invasion recognition particle tree, and detects mixed diverse attack groups in complex network environment according to error estimation. Experimental results show that the algorithm can effectively improve the accuracy of detection in complex network environment. atures;