The Application Of Roi In Remote Sensing Image Coding And Transmission

With the rapid development of the internet technology, the image has become an indispensable part of people’s lives. Because of the high resolution and intuitive, the remote sensing image is widely used in industry, agriculture, military. JPEG2000as a new generation of still image compression standard not only can improve the deficiencies and the compression ratio, but also have a lot of excellent new properties. For the remote sensing image, JPEG2000can be efficiently compressed, while using progressive transmission and region of interest coding to meet the scalability requirements of remote sensing image.Firstly, this study proposes the remote sensing image compression and transmission system, and it mainly includes two parts:regions of interest extraction and coding. In this paper, the region of interest is determined by the high frequency information of wavelet transform. The method proposed in this paper can extract ROI automatically, and the shape and number of region is not restricted. The method mainly designed and improved at the wavelet transform and the post processing which is after the mask’s extraction. According to the limitation of the wavelet basis in the JPEG2000on the shape, the wavelet basis cannot fit the extraction method proposed in this paper. So using shape adaptive wavelet transform make replaced. The innovation of this paper is put forward the post processing of ROI mask. The post processing could adapt to Tie-1layer coding at the same time. The parameters of post processing procedure can be adjusted to control the effect on the subband of wavelet transform, so as to control the transmission sequence of ROI. Human visual characteristics is also a need to be considered in the post processing, in order to achieve the lowest data for the best compression effect. The experimental data shows that the system can extract the region of interest automatically, while the experimental results meet the requirements of remote sensing image on transmission. The quality of image restored of this system is higher than that of JPEG2000system in the same amount of data, when the data are not fully received.