Recognition And Classification Of Liquid Substances In High-speed Inspection Equipment

In recent years,people’s living standards have become increasingly higher,and transportation has become more and more convenient.In such a context,the increasing number of people are more willing to go out and are more pleasure to spend,travel,and shop.On the one hand,these changes and activities are contributing to social development.On the other hand,they bring a lot of security risks.In order to ensure the personal safety of people,especially in some crowded areas,such as airports,railway stations,the security inspection process is very essential and important.At present,X-ray security inspection equipment are used commonly in the airports,the railway stations and other places.With the high detection efficiency and good recognition effect,X-ray security inspection equipment can basically meet the detection requirements of these places.However,during the process of detection,manual real-time observation is needed.Moreover,the current technical means can only tell whether the matter can be carried or not from the color and image displayed in the abovementioned equipment.With regard to liquidlike matters,using this method is difficult to determine whether the matter is allowed to be carried or not.Facing such an embarrassing problem,so in the railway station,the subway and other places,it is even necessary to open the lid of bottles for identification.Even worse,it is clearly stipulated that any liquid is not allowed to be carried at airports.Those rules and requirement has caused a certain bad impact on people’s travel,while making a lot of waste.This paper focuses on the extraction of image information and classification of common liquid items.In terms of the extraction of image information:the image generated by X-ray penetration usually contains some interference information due to device vibration and other reasons.For this,a new wavelet threshold function is reconstructed,and the wavelet analysis is combined with contourlet analysis to improve the denoising effect;Secondly,for the relationship between the gray level and the ray attenuation of the X-ray transillumination image,the true gray level between the overlapping substances is analyzed.on the basic of this,a new method for removing the influence of high-density containers on the real gray level extraction of liquids is proposed.In the classification of common liquid substances:Through experimental methods,X-ray fluoroscopy experiments on acetone,kerosene,water,and hydrochloric acid are conducted,the high and low energy gray level in thin plastic containers are collected,the curve fitting of the data collected is carried out through the Origin software,and the characteristic curves of four liquid substances are finally drawn,which is used as a criterion for determining the type of liquid.At the end of the paper,the same linear scanning experiments are performed on the four liquids using glass and iron containers,the analysis of the removal of the influence of the vessel wall on the data collected is carried out,thus,the feasibility of the proposed new method for removing the influence of containers is verified.