Research On Automatic Image Analysis Technology Of Network Equipment Based On Eigenvalue

With the advent of the information age, network security problem also receives more and more attention, network equipment as the carrier undertakes interconnect different network, enable each node in the network free and distribute right transfer data and information to the network corresponding address. Network equipment security is an important means to ensure network security, so the analysis of network equipment security is the key to network security research. In the structure of the network equipment, the firmware stores all the information of the network equipment for running normally, so the research on the firmware of the network equipment is an important means and method to study the security of the network equipment. In in the research of network device firmware, first of all to get the firmware image content, only to get the correct configuration information we could use the correct strategy to reverse network device firmware, so the analysis of network device firmware image is an important part of the research on security of network equipment.Based on the analysis and study of current network equipment image technology, the status of development strategy, it is found that, at present for network equipment image analytical research mainly concentrated on the study of firmware images as a function, by extracting the image of various types of function to restore the purpose and function of the mirror, rarely take into account the mirror configuration information acquisition. In addition, the analysis of mirror function usually used in reverse assembly technology, high technical difficulty and the disassembly of the accuracy can not be guaranteed, there is a huge number network equipments, so we can not use this method as a general research to network equipment.In view of the above situation, in this paper, combined with the network device firmware image file data form we put forward a kind of automatic matching image resolution strategy based on feature value, the strategy mainly consists of two parts: for the different types of network devices of mirror type recognition strategy and for specific types of network equipment firmware image parsing strategies. For different types of network devices is separate various kinds of various types of image, access to the mirror type of network equipment, and for certain types of network equipment firmware image analysis refers to distinguish mirror type of firmware images for further analytical work, to obtain more detailed configuration information. For image classification of different types of network devices using network equipment image in characterizing the mirror type characteristic field is used as the basis for distinguishing, through the analysis of image format plugin and means of verification of the reverse access to effective field feature and the Java Script object notation(JSON) to store data, and then use a discriminant logic matching recognition of the characteristic data of the analysis of the network equipment data in JSON format, acquiring a mirror image type. For certain types of network equipment firmware image analysis through further analysis, extract details of the equipment characteristic field and in accordance with the characteristics of a field of characteristic data were data modeling, well structured feature database is constructed, then according to the analysis of the requirements, to treat analysis of network equipment image data separation, reorganization and match up with the characteristics of the database data, thus the detailed identification of the network device mirror related attributes and information.According to the core idea of each strategy, this paper use SQLite database and python language realize the image analytical automation tools. After verification, the method in this paper can accurately complete the network device firmware image recognition and analysis.