Indoor Security Millimeter Wave Radiation Characteristic Analysis And Image Feature Extraction

Since the increase of terrorist incidents and the deterioration of national security issues, the world begin to pay attention to the public safety problems, at the same time, they focus on improving the accuracy and effectiveness of security detection system. The metal security doors, X-ray detectors and infrared detectors which were widely used in the early public security systems have their respective shortcomings, while the appearance of passive Millimeter-wave(MMW) system has made up the shortcomings. MMW has strong penetration, high precision and the ability to work all-time, all-weather. On the other hand, the passive MMW system does not radiate electromagnetic waves, and there will be no damage to human or other objects.In our country, due to the limitation of technology and equipment, the passive MMW security system is still in the initial stage. So this paper focus on the analysis and research of MMW system, including basic theory of MMW, radiation characteristics of concealed objects hidden under clothing, passive MMW imaging systems and image processing, the summary as follows:(1)Introduce the development history of the security system, expounding the advantages of MMW imaging applied to security systems and the research status both in domestic and overseas.(2)Describe the basic theory of MMW radiation, the relationship between antenna temperature and the radiometer power. The full power radiometer and comparative radiometer have also been introduced about their technical indicators, advantages and disadvantages.(3)Use two methods to analyze the radiation characteristics of the concealed objects under human body. From the perspective of radiation measurement, we built the analysis model of antenna temperature contrast between the concealed object and the human body, then discussed the relationship between antenna temperature contrast and clothing transmissivity under the circumstances of indoor and outdoor. On the other hand, from the perspective of multi-layered medium, the surface brightness temperature model of the concealed objects was built. Then calculated the coherent reflectivity by using the equivalent transmission line and analyzed the relationship between the clothing layer thickness and coherent reflectivity.(4)Simulate the indoor security imaging process by using the laboratory MMW system, and get the pseudo-color image of metal object. In order to process the image, the basic theory of morphology algorithm is introduced, so as the applications in image denoising, enhancement and boundary extraction. In this paper, we propose a combination morphological filter to smooth and filter the image.(5)Segment the denoised image so as to extract the features of human body and concealed objects. Firstly, several common edge extraction methods and segmentation techniques are discussed. Then watershed algorithm is proposed to extract edge feature from the denoised image. Compared with the edge detection differential operators, we can figure out the pros and cons of the watershed algorithm.(6)Summarize all the work in this paper, analyze the deficiencies and improvements.