Study On Rapid Preprocessing Of Massive MOST Images Based On Cluster

Humans have long been eager to know the brain,protect the brain and even create the brain,but we still know very little about the brain.Structure is the basis for determining function.To understand the brain,we must first understand its structure.The neural circuits formed by the interconnection of neurons are responsible for the information transmission and processing in the brain,thereby exerting the functions of the brain.So parsing the neural circuit becomes the basis for our understanding of the brain.Due to the particularity of the structure of neurons,to study the brain,we must obtain imaging data of the fine structure of the single neuron level in the whole brain.In the past ten years,scientists from all over the world have developed a variety of whole-brain high-resolution imaging systems represented by Micro-optical section tomography systems.However,the problem with such imaging methods is that the image data sets acquired by the system are often terabytes.The existing processing software cannot quickly complete the preprocessing of the data set.In this paper,I have studied about a rapid preprocessing method for massive image data collected by Micro-optical section tomography.I have improved the defects in the established image preprocessing scheme.Three new image processing applications were designed through modular development,which are the new fluorescence micro-optical section tomography system image stitching preprocessing application,massive image batch processing application,fast acquisition of high resolution XZ/YZ plane projection map application.These three applications have been added to the original image preprocessing scheme to make it more perfect.The original massive image preprocessing scheme can support the preprocessing of a new type of data by using image mosaic pre-processing application.Massive image batch processing applications and fast acquisition of high-resolution XZ/YZ planar projection map applications can achieve faster processing results than the original processing scheme.The three image preprocessing applications are designed to use MPI and OpenMP two-level parallel acceleration to quickly process massive image datasets with high resolution on the CPU cluster.In this paper,an image stitching preprocessing scheme is established for the new system,which can quickly complete the preprocessing of the original dataset collected by the system.In addition,a new processing scheme for mass image batch processing and high-resolution XZ/YZ plane projection map function is proposed,which can obtain processing results faster than the original processing scheme.The three solutions proposed in this paper are designed as independent applications,which can quickly process massive image data on high-performance server clusters.