Remote Sensing Images Based On Wavelet Transform To Compress The Rapid Implementation Of The System

Image compression is one of the key techniques in today's digital image-processing field. It has wide applications. The considerable body of research over the last two decades has shown the significance of this technique. The early research on image compression mainlysfocused on the discrete cosine transform(DCT). While the recent research mostly focus on multi-resolution techniques, particularly the wavelet transform.Image compression plays a significant role in the field of remote sensing technology. Considering many important characteristics of remote sensing images, such as high resolution, little correlation, low redundancy, rich texture, etc., we have proposed a WT-based, fast encoding image compression system for remote sensing images.At first, the theory basis, characters and property of wavelet transform are analyzed from the view of mathematical transformation , the practical steps and their flow charts are provided in this paper. After a remote sensing image ,has been decomposed by wavelet transform, low frequency sub-image and high frequency sub-images can be quantified and encoded with different techniques that match the human visual , physiological and psychological characteristics. In essence, the coding algorithm involves performing RIDPCM coding method on the wavelet coefficients of low frequency sub-image and performing adaptive quantization and arithmetic entropy coding method on the wavelet coefficients of high frequency sub-image. In order to accelerate procedure's operation speed, the key codes of the software in image compression system are designed in assembler language. Then the procedure is optimized by MMX instruction and put at the mode ofRing 0 level. Therefore, the image compression speed and the stability of the compression system are enhanced.With this compression system, the compression quality can meet the compression demands of remote sensing images when the compression ratio reaches 4:1 or 8:1, but the compression time this system has been increased by an order of magnitude, compared with that of JPEG 2000 or JPEG.Wavelet transform and its application are hot points in recent domestic and international studies. The study method in this paper has some theory direction and practical application value for various fields of image data compression processing using wavelet transform.