Research On The Method Of Interference Sources Identification For UV Detection

For simple signal processing, low false alarm, no scanning and cooling, passivedetection, etc., UV warning system has been widely researched and applied, the amountof equipment is the largest one of the missile warning system by now. Domesticresearch on UV warning technology has also been carried out in the last two decades,compared to abroad, and the main deficiency is high false alarm rate. In this paper, forUV warning images susceptible to noise interference and the system vulnerable toground emitter interference leading to the problem of high rates of false alarm in thesystem, research work of the target identification and interference sources recognition ofthe UV warning images were carried out.Principle in each link of UV warning system was analyzed. Blind UV detection isdifferent from infrared image processing, ultraviolet image background processing isrelatively easy, but the image is seriously affected by random noise. Random noise andobjectives in a single frame both with high-frequency，high grayscale characteristics,they can not be distinguished within a single frame. Noise processing methods basedon image sequences, such as the time-domain recursive, neighboring region verdict,pipeline filter and so on were studied, the UV image sequences containing a movingtarget was processed, and the processing result was detailed analyzed.Social activities cause a complex clutter ultraviolet environment. Various types ofarc discharge light sources in industrial production which produce strong UV radiationare the mean false alarm sources to UV warning system. On the basis of simulationresearch on characteristics of the missile target and experimental measurement of theradiation characteristics of interference sources, interference source identificationmethods based on target motion characteristics and time-domain UV radiationcharacteristics were studied, specific algorithms were given, using time domaincharacteristics identification method processing ultraviolet image sequences. The resultsshow that the identification method based on time-domain characteristics can effectivelydistinguish the missile target and UV lights, welding and other sources of interference.