Research And Implementation Of Algorithm For Detection Of Dangerous Goods

In recent years, the international situation is turbulent, terrorist attacks occurred frequently. Public safety problems have become very serious, and also, more and more endangering public safety events occur in the bus. Criminals often use gasoline, knives or other tools to harm people, it causes extremely bad social influence. Therefore, it is very important to study an effective blind source separation algorithm for the detection of complex mixed signal.Blind source separation(BSS) studies the premise of even source signal and its hybrid mode do not know, only by some statistical characteristics of the source signal, and then according to the observed signals(mixed signals) to recover the source signal. Independent component analysis(ICA) is a special algorithm for blind source separation, which is based on the independent of source signal. Fast ICA algorithm is used for maximize the signal of non-Gauss, the algorithm uses Newton's method to calculate the value of multiple sampling data, each step of the iterative calculation can separate an independent components from the mixed signals, which belongs to the ICA algorithm of a kind of fast calculation method.In this paper, we analyze the principle of microwave detection of dangerous goods, and the function of detecting system is implemented which based on the principle of detection. It based on the dielectric constant of the gasoline, ethanol and other common dangerous goods to detect S-parameters of dangerous goods with remote and non-contact, and it analysis the S-parameters. Finally it achieves investigation of dangerous goods with using centimeter wave upon remote and non-contact methods. Then the algorithm is studied, and two kinds of algorithms are used for different situations: the accumulation algorithm and the FastICA algorithm. We need to determine what dangerous goods is when only detected one dangerous goods. In this paper, the cumulative sum and the accumulation algorithm are introduced respectively. Finally, according to the result determines that the accumulation algorithm is more applicable in the detection of dangerous goods. If simultaneously detect a variety of dangerous signals, and the signals are an observed signals, so we need to use Fast ICA algorithm to separate the blind source and get the independent signal of different dangerous goods. In the paper, we use the water and air as a reference, by comparison, it is found that the separation signal obtained by water is similar to source signal, which is closer to the air.In order to implement the algorithm for the detection of dangerous goods, a software system for the detection of dangerous goods is designed. It is composed of the PC part and the lower computer part. When developing a software system for the detection of dangerous goods, I compare with some existing interface development software and tools, and finally choice the Qt framework. In the aspect of communication, the high reliability of the TCP is used for data transmission, it uses the TCP client/server mode. The lower computer request to connect the PC and the connection is successful, the lower computer can start sending information to PC and PC can receive data from the lower computer. When the PC receives the detected information of dangerous goods, it will be displayed the images of dangerous goods and description of the picture on the interface, at the same time the PC will have a warning sound. We can also save screenshot of the received information.In the process of the research and implementation of the dangerous goods detection algorithm, the algorithm and the code are constantly improved and debugged to correct various detail errors. It implements to distinguish the dangerous goods, and through the dangerous goods detection software system to implement the alarm. The stability of communication, the function of the upper computer and the lower computer has basically reached the design requirements.