Search

Rapeseed is one of the important oilseed crop species, and is worldwide the most important economically. The light seed colour in rapeseed is associated with improved oil, protein and fibre contents. But there are not reliable and efficient methods to measure seed colour, especially the single seed colour. Two transformations of RGB (red, green, and blue) colour space were used for two seed colour recognition methods, i.e., HSV (hue, saturation, and value) and nine colour model (NCM). Using these two colour space transformations, the performance of the common method on rapeseed colour recognition was compared with the major colour method. The common method obtained the colour recognition accuracy of 83.96% in single seed recognition and 92.72% in sample recognition. The major colour method obtained the colour recognition accuracy of 98.91% in single seed recognition and 100% in sample recognition. The major colour method combined with HSV and NCM colour space transformation proved to be a good approach for seed colour recognition of rapeseed using machine vision.

At intelligent methods for information audit system is hot spot in the field of network security, and application of pattern recognition and data mining in information audit system is world widely concerned and worldwide studying. As an important method of pattern recognition, fuzzy neural network has the capability of self-organization, self-learning and generalization. Application of fuzzy neural network in information audit system can not only identify the known badness information, but also can detect the new badness information and abnormal event.

Since the method of using username and password and the traditional access-control list mechanism can no more keep up with the further development of e-government, we realized a distributed-access-control model to settle this problem. This model divides the security management into two levels: First is global security management level which is responsible for harmonizing and controlling the security management of all the control objects of the e-government network; Second is local security management level which is in charge of the security management of local departments or objects inside local project group. By recompiling of the kernel of FreeBSD 5.0 version RC2, we have realized the access control model described above.

The paper proposes a new network intrusion detection system based on fuzzy neural network by redesigning the intrusion detection system's architecture and arithmetic. In order to overcome the difficulty of specifying the membership function of rules depending on experiences of experts in multi-dimension space., neural network is introduced to distinguish non-linearly input/output characteristics of complex system and to generate rule sets and membership functions automatically. The new architecture adopts the network processor to collect and analyse the data in the low layer of network., and to establish a prototype system. This system demonstrated in this experiment appears to be better intrusion detection ability, moreover, which is able to detect unknown attack and plays down false alarms.

With the deeply research on intrusion detection techniques and the widely use of intrusion detection products, the study of evaluation techniques of intrusion detection systems became important. In the paper, relative works of the evaluation of intrusion detection systems was introduced. The primary aspects of intrusion detection systems in an evaluation were discussed. An evaluation system for intrusion detection systems was proposed. In the system, a supervisor module controls the whole system. The supervisor module schedules the traffic control module and the attack emulation module. The data in the evaluation environment are recorded and input to the evaluation module. The functions such as emulation of network traffic and host usage, emulation of attacks and evaluation report generation are implemented.