Research On Filtering Processing Technology Of Speckle Pattern Interferometry Fringes

Electron speckle interference images are a special type of image.When the diffuse surface is illuminated by laser light,there are randomly distributed bright and dark spots in the space,which was called speckles.The use of speckle will change with the object's deformation or movement characteristics,than can be highly accurate detection of the surface displacement of the object,which is electronic speckle interferometry technology.The noise in the speckle image can reduce the quality of the image to a large extent,and can not ensure the accuracy of the final extraction results.Image filtering methods and algorithms are diverse,the more traditional spatial domain filtering in the smooth noise,but also easy to make the image details and edge inforrmation is damaged.Although the frequency domain filtering algorithm is superior to the space domain to a certain extent,there are still some problems such as complicated operation and slow operation speed.In terms of the noise characteristics of the speckle image itself,the number of speckle particles is large and disorderly.Therefore,it is necessary to start from the noise characteristics of speckle images using special image processing methods in order to more accurately extract the effective information in the image.In thesis,an improved SVM-PCNN filtering algorithm is proposed.Combined with the advantages and features of translational invariant shear wave SVM and PCNN,the algorithm combines the speckle features with the details of the algorithm Compared with the experimental results of other filtering algorithms,it is proved that this filtering algorithm can better filter out the speckle particle noise and protect the edge of the image detail significantly.Using this algorithm to filter the image of the electronic speckle fringe fringes,the image has the advantages of high contrast and clear fringe skeleton.This thesis's research work can be summarized as follows:(1)How to obtain speckle fringe patterns by using electronic speckle interferometry,and the characteristics of the whole distribution of fringe patterns and speckle particle noise.(2)The speckle fringe image is processed in frequency domain and time domain,including wavelet transform,homomorphic filtering and new fourth-order partial differential equation model,and the results after filtering are given.(3)Stripe refinement and peak signal ratio etc.can be used as an index to evaluate the effect of speckle interference fringe image filtering.Starting from these aspects,we analyze and discuss the filter effects obtained by applying various filtering algorithms.Therefore,it can be argued from the objective indexes that the filtering algorithm proposed in this thesis can achieve better filtering effect and is a feasible speckle interference fringe image filtering algorithm.