Design Of Image Content Filtering Firewall Based On Web Page Logo

With the rapid development of the network and the fast increase of the user of Internet, the security of network has been become a more important problem for the safety of the country and the society. Many kinds of information, such as bad and reactionary message, the security secret of country and so on, is more and more disseminated by Internet, which is strided the region or country and opened to the world. In order to solve this problem, it is important to monitor and filter the information of network except for punishing these manners about commit crime and violate law by law.The firewall is the most widely useful technology, it can improve the problem who can't monitor information well. Nowadays, most of kinds of firewall which filter based on IP address, port number, URL, text information and so on have not enough ability to filter some new things. With the development of multimedia technology and enlargement of network bandwidth, the proportion of image and video has been largely increased in Internet. However, the present firewall that is used to filter image content has lower ability which is about real time and accuracy filtering, and it can't be satisfied with the demand which is fast growing.Firstly, this article begun with the present situation of network security, and it discussed the key point of network security. Then the firewall has been led out. Secondly, it introduced the history of the development of firewall, the definition of firewall and the common classification, then a technology about image content filtering firewall based on web page logo has been brought, and it can greatly solve the problem of firewall which has not enough ability to filter the image content. Thirdly, the article adopted experimentation to contrast over and over again, after threshold is reasonably decided, it designed and achieved a image content filtering firewall which is based on web page logo, and it synthesize many kinds of algorithms to contrast the similarity among every logo, then it can not only solve the lagging problem which is brought by IP filtering etc. method but also lower the miscarriage of justice probability which is brought by text analyze. At the same time, the filtering velocity problem is influenced by analyzing images that contain much pixel can be avoided. Finally, it compared with interrelated technology which is about image content with higher accuracy and lower erroneous judgment, and it can raise its filtering ability and filtering velocity greatly.