Research On Microscopic Image Acquisition And 3D Visualization Of Wheat Grain Based On Micro-CT

Food security is an important prerequisite for social harmony,economic stability.The traditional grain pest detection method has many shortcomings,such as too strong subjectivity,difficult to find hidden pests.In recent years.researches find various new detection technology on the detection of grain pests deeply,but automatic detection of inner eggs and young larvae in grains could not be realized.Micro-CT imaging on the grain inside the egg and the young larvae by scanning imaging system can obtain high resolution stereo information of grain and analyze can change the internal micro-structure,the detailed analysis of three-dimensional visualization can applied on single grain's inner structure,moreover,it can solve the early automatic detection of pests in grains.The main contents of this paper conclude there parts: the acquisition of wheat grain sample and Micro-CT image collection,the Micro-CT image reconstruction of wheat grain and the three-dimensional visualization of wheat grain.Construction of a multi-grain fixate with long axis orientation and keep the grain at the center of the scan in the right position.Using Micro-CT Image acquisition system,When the anodic voltage of X ray is 55,the current is 55 A,the filter is AL,and the integral time is 200 ms,in the 3rd day,9th day,17 th day,22 rd day and 28 th day then collect the egg stage,low instar larval stage,old larval stage,pupa stage and adult stage images,respectively.The grayscale cube with grain thickness as the same position in 1-6mm stereoscopic image data was selected for analysis.The results showed that with the increase of wheat grain thickness,the gray value of gray cube decreased and the contrast of wheat grain gradually decreased.By the projection image of the long axis of wheat grains and the rotary table angle at 0 °,45 ° and 90 °,analysis of the relationship between rotation angle and imaging effect.The results showed that with the increase of the angle between the long axis of wheat grain and the rotating table,the overall contrast of wheat grain would also increase.Taking the Shepp-Logan head model as the object,using filtering back projection method?algebraic iterative reconstruction algorithm?joint algebraic reconstruction algorithm and statistical iterative reconstruction algorithm to realize the simulation and reconstruction of the header model.In the case of incomplete projection data,reconstruction of FDK algorithm and FDK algorithm and MLEM algorithm,The results show that the image quality of SART algorithm is best and it can clearly reflect inner information in grains,the iterative convergence speed of MLEM algorithm is fastest and image quality is better than FDK algorithm,but the result of reconstruction is worse than SART algorithm.Under the fully projected data,FDK algorithm has the best effect on image reconstruction.The mixed filter function is used to reconstruct the projection data of wheat grain in FDK algorithm.The results show that the mixed filter function method can obtain higher contrast and less noise slice image of wheat grain.After segmentation,morphological operation,filtering and other processing of wheat grain slice,2D slice image of wheat grain suitable for 3D visualization is obtained.Gaussian low pass filter is used to smooth 3D data network and Phone model is used to optimize lighting setting,the visualization based on MC algorithm is realized.The results show that the visualization of MC algorithm on MATLAB platform can restore the inner structure of wheat grain well.On the platform of Mimics,3D visualization of wheat kernels is realized,and the interference factors such as internal cavity and partial cracks are removed by segmentation and filling.The results show that Mimics platform can directly process the inner information of wheat grain and get the effective feature.