Size Detection And Accounting Of Small Particls In Aero Oil

Along with the aviation industry rapid development, application of thehydraulic system is more and more widespread on airplane. The hydraulicequipment operational reliability and the service life has the direct relations withthe hydraulic system pollution condition. According to the statistical data indicatedthat about 60%-80% hydraulic system breakdown is caused by the fat liquorpollution, of which resulted by solid particle pollutant occupies 60%-70%. A fatliquor pollution examination is mainly determines the solid pollution pellet size andthe grain-size distribution in the fat liquor. The edge examination is the foundationof pollution pellet size classification statistics, and its examination result directlyaffects to the judgment of fat liquor pollution condition. With the rapiddevelopment of modern micro photo-electricity imagery and the computer pictureprocessing technology, uniting the computer picture processing technology and fatliquor pollution examination has already become popular topic of the aviationhydraulic system attrition diagnosis and the fat liquor pollution examination.Only to measure the appropriate edge of the obtained particle picture, coulddraw pollution pellet size, classify it and confirm the class of pollution, then takethe corresponding counter-measure. So edge detection of pollute pellet pictureplays a non-negligible role in fat liquor examination of pollution index. This articleelaborated the molecule examination system principle and counts the method basedon in the digital image processing aviation fat liquor. To gathers the pollution pelletcarries on the edge examination through the model tradition and the modern edgeexamination operator to, the use best threshold value iteration algorithm withdrawsthe pellet image edge information, carries on the imagery processing programmingthrough VC and the MATLAB higher order language, has completed to the aviationfat liquor molecule image examination function.The computer image processing is the procedure that some given pictures aretransformed in order to get sharp image. First, the study of object must be splitfrom the image, then feature extraction is carried on. The characteristic of image isthe property which can be as mark in image field. It has many shapes. But it can bedivided image statistical and vision property generally. The characteristic of imagestatistical is some artificial property and could be get through by transformation,for instance, image histogram, moment, and frequency spectrum, etc.. The visionproperty is the nature property which can be felt directly by human vision, forinstance, the area luminance, texture and profile, etc.. That the procedure of whichthe image is decomposed into a series of meaningful goal or area by these twokinds of properties is defined as image segmentation.The computer image processing is the procedure that some given pictures aretransformed in order to get sharp image. First, the study of object must be splitfrom the image, then feature extraction is carried on. The characteristic of image isthe property which can be as mark in image field. It has many shapes. But it can bedivided image statistical and vision property generally. The characteristic of imagestatistical is some artificial property and could be get through by transformation,for instance, image histogram, moment, and frequency spectrum, etc.. The visionproperty is the nature property which can be felt directly by human vision, forinstance, the area luminance, texture and profile, etc. That the procedure of whichthe image is decomposed into a series of meaningful goal or area by these twokinds of properties is defined as image segmentation.The essence of edge detection is to adopt a certain algorithm to draw the boundedline among the particle target and fluid background in the picture. Because the surfaceroughness of pollution pellet is different, the reflected component of the light is alsodifferent. So adopting the same collected pollution pellet picture to draw the edge withdifferent operator, the result will be different to some extent. There are many edgedetection algorithms at present. And several practical and classical detection operator andbest threshold value operator are analyzed detailedly in the article. Through detecting theparticle picture that is gathered actually, they have been verified.This article designs a detection system which is comprehensive of laser technology,CCD and picture processing of computer technology. The light beam sent out by thelaser instrument produces the parallel light beam through the collimation and expansionbeam system. The parallel light beam shines to the sample oil which is to be detected.Through finely tuning sample pool horizontal location system, it is lighted evenly andsufficiently. The collected data of sample object picture taken by CCD camera via imagecollection card is processed and analyzed with computer image processing software.Statistics result of particle diameter size is then classified. And the grade of this fat liquordustiness is demonstrated on the monitor screen. Meanwhile, the test result can also beprinted and the test data can be stored.Demarcates through the system with the modelling, molecule image severalalgorithms has carried on the recognition to the sample fat liquor in. The resultindicated that, in the article the pretreatment method which uses to the image thepellet image can effectively divide to the image in. But the test result still had thecertain error. Through to tests the result and the system software and hardwareanalysis research, analyzed this test system to have the erroneous reason.