Research On UAV Remote Sensing Image Segmentation Based On Hadoop Cloud Platform

With the continuous development of UAV technology,the application of UAV remote sensing is becoming more and m ore extensive and gradually penetrates into all works of life,and th e resulting UAV remote sensing images are also in creasing.In the face of massive UAV remote sensing images,how to quickly and effectively segment the image and extract the ef fective image information has become a key issue in the field of UA V remote sensing application.Aim ing at the pr oblem that the current image segm entation method is slow and ineffective for remote sensing image segmentation,this paper uses the open source cloud computing platform Ha doop to study the remote sensing i mage segmentation method of UA V,and designs the MeanShift algorithm of UA V remote sensing image segmentation based on Hadoop cloud platform.Firstly,the im age processing m odel of Hadoop is designed,custom design of MapReduce image processing programm ing model,and how the Hadoop platform reads images to achieve image reading quickly and efficiently,to design image serialization,file segmentation and file output type,to build th e Hadoop cluster and test the stability of the cluster operation.Then the MeanShift algorithm for Hadoop platform is designed,and the image preprocessing algorithm transferred by Hadoop platform is to preprocess the im age and to achieve the MeanShift im age segmentation algorithm,which is based on the Hadoop platform in the OpenCV.The experimental results show that under the prem ise of ensuring the consistency of the segm entation parameters and processing conditions,the efficiency of UAV image segmentation based on the Hadoop platform has been greatly improved,Compared with the single MA TLAB image segmentation,the concurrency,effectiveness,computing and storage capacity are greatly enhanced.