Research on remote sensing Digital image processing method based on Hadoop

Research on remote sensing Digital image processing method based on Hadoop

Northeast Normal University Zhou

Based on the Hadoop cloud computing system, this paper mainly uses the parallel programming framework MapReduce to realize the enhancement processing of remote sensing digital image and clustering the enhanced image, and compares it with PC serial processing. In view of the low overall brightness and poor visual effect of remote sensing digital image, the traditional image enhancement method can not achieve the visual interpretation of the human eye comfort and the problem of subsequent processing, and this paper improves the method based on the nondestructive enhancement. The results show that the effective image area of the original image can be enhanced by the invalid pixel, the experimental result shows that the method achieves a good visual effect. In this paper, a parallel processing model of "waterfall" is proposed for the characteristics of large amount of remote sensing digital image, and the model is realized by MapReduce parallel programming in the environment of Hadoop cloud computing platform. It not only solves the problem of mapreduce directly processing the commonly used image format, but also accelerates the processing speed of remote sensing digital image. At the end of this paper, the image clustering for the enhanced processing of remote sensing digital images is carried out, and the experimental results show that the improved image enhancement method presented in this paper has a good effect on clustering, and the application of K-means algorithm in MapReduce is improved in remote sensing image. The process of selecting the pixel of clustering and the segmentation selection of the initial cluster center is increased.

Research on remote sensing Digital image processing method based on Hadoop