Optical Flow Displacement Measurement Based On Fringe Frequency Extraction

With the continuous development of modern technology and industry,optical displacement measurement plays a pivotal role in modern industrial detection,military communication,three-dimensional sensing,biomedical engineering and other fields with its advantages of non-contact,high precision,simple operation,full field rapid measurement and low cost.The concept of optical flow was first proposed in 1950.It refers to the instantaneous speed of pixel movement brought by object movement on the two-dimensional imaging plane,reflects the structure information and dynamic behavior information of object surface,and can be used to approximate the real sports field,so as to study the relationship between image gray level change and object structure and movement.Its advantage is that it can detect independent moving targets and extract complete information without predicting scene information.Therefore,it has broad prospects in image processing,computer vision,target recognition and tracking,and can be used for visual positioning,pattern recognition,stereo matching,motion detection and other aspects.The essence of optical flow research is to solve the motion vector of pixels in two adjacent frames of gray scale.In the projection fringe image,the fringe optical flow field formed by pixel movement caused by objects placed can be used for deformation measurement to obtain in-plane displacement field,and then the relationship between optical flow and out-of-plane displacement can be derived to restore the three-dimensional morphology of objects and obtain the coordinates of space points on the surface of objects.In the calculation of fringe optical flow,the key to achieve out-of-plane displacement measurement through fringe deformation is the extraction of fringe frequency at each point of the image.Therefore,this paper proposes a local fringe frequency extraction method based on the window Fourier ridge method.By improving the original algorithm,the influence of interference items is eliminated,and the algorithm is applied to the measurement of optical flow displacement.The measurement accuracy can be improved.The main research contents of this paper are as follows:1.The concept and basic classification of image matching algorithm are introduced,and the theoretical basis and advantages and disadvantages of the algorithm to achieve displacement measurement are expounded.Two methods of measuring the degree of image matching by correlation calculation are introduced,namely,the normalized correlation image matching algorithm based on spatial domain and the phase correlation image matching algorithm based on frequency domain.The principle derivation and detection result analysis of the two algorithms are given.Through simulation,it is verified that the two methods can realize the similarity detection and the solution of the whole pixel level displacement in the opposite side,and the factors that affect the minimum template size required for matching are explored.Finally,the optical flow measurement technology is proposed in view of the shortcomings of the two correlation matching algorithms in displacement measurement.2.The concept hypothesis of optical flow is introduced,and the unique solution of optical flow components u and v can be obtained only by combining the basic equation of optical flow with other constraints.The solution idea of optical flow method under three different constraints is introduced.The principle of optical flow three-dimensional displacement measurement technology is deduced.For the two projected fringe images before and after the deformation modulated by the subject plane,the in-plane displacement field can be obtained by calculating the fringe optical flow field when the time interval is 1.The three-dimensional surface distribution of the object can be obtained by using the mathematical relationship between optical flow and out-of-plane displacement,and the fringe frequency is an important parameter.The simulation results compare the relative errors of the theoretical values and the calculated values of the optical flow method,and prove the advantages of the optical flow method in the measurement of in-plane and out-of-plane displacement,which is not limited by the choice of projection pattern.3.An improved fringe frequency extraction method is proposed for optical flow displacement measurement.Firstly,the technical principle of extracting local fringe frequency by traditional window Fourier ridge method is given,and the necessity of adding window to image is expounded.Through simulation,the shortcomings of this method in extracting frequency of complex fringe image are proved.The obtained spectral amplitude graph is analyzed,and then the improved normalized spectral method is proposed.The simulation results show that the proposed method can eliminate the interference caused by the frequency(0,0)when the frequency amplitude is larger than the real frequency amplitude,and the precision of local fringe frequency extraction is improved.Finally,the algorithm is applied to the experimental fringe graph obtained by the Michelson interferometer to obtain the frequency distribution.The phase information of out-of-plane displacement can be measured effectively by combining the improved fringe frequency extraction algorithm with the optical flow algorithm.