Realization Of Moire Image Processing Method Based On Hartley Transform

The use of Fourier transform to deal with moire images is a common method for analyzing three-dimensional information on the surface of objects.Since the Fourier transform is a complex number,the larger amount of computation in the calculation process limits the speed of processing the moire image.In view of this situation,in this paper,a moire image processing method based on Hartley transform is adopted,Hartley transform is a harmonic transformation similar to Fourier transform,has inherent consistency.Hartley transformation has the advantage that the process is a real number operation and that the positive and negative transforms are consistent.In this paper,we use the Hartley transform to process moire image,including positive transform,filter,frequency shift,inverse transform,defect images and phase images are obtained from moire images,and compared the processing of moire image processing based on Fourier transform,then,based on the problem of large phase jump point noise in one-dimensional case,the moire image processing algorithm based on two-dimensional Hartley transform is adopted,which weakens the noise of the jump point and obtains better results.And then analyzed because the defect is located in the edge of the moire image,affecting the other boundaries of the moire curve,we use the stacking method to eliminate this effect.In addition to accelerating the detection from the perspective of the algorithm,you can also from the software design point of view can also be used multi-core parallel way to speed up.This paper evaluates the time complexity and spatial complexity required by Hartley transform and Fourier transform,and proves the superiority of Hartley transform in many aspects.Because of the irrelevance between the moire images,you can take the data in parallel to speed up the processing.In this paper,we discuss the relationship between the image and the number of threads in the multi-thread environment.Finally,the detection rate based on the Hartley transform is based on the Fourier transform detection speed of 3-4 times.