Application And Optimization Of EZC-DCT Compression Algorithm In Terrain Rendering

In recent years,large-scale terrain real-time rendering technology has played an important role in more and more fields.At the same time,more and more terrain data have a great impact on the data transmission process and rendering rate.However,most of the civil equipment performance and network speed are general,so,even the basic real-time requirements are difficult to meet.In many applications,the demand for real time is far greater than the requirement for terrain accuracy.Therefore,the algorithm,which focuses on the rate of terrain data compression,terrain rendering speed and the general performance equipment,will be the key to solving the above problems.As an excellent performance image compression algorithm,EZC-DCT can have the advantages of both the multi-resolution characteristics of the wavelet transform and the simple operation of the DCT transform,so it can guarantee the lower load of the equipment operation and the higher compression ratio of the data compression.Based on this,this paper makes a thorough study of the existing EZC-DCT algorithm and some related image compression algorithms.At the same time,combining with the GPU related technology and facing the general performance equipment,a EZC-DCT terrain compression algorithm suitable for real-time terrain rendering and its optimization scheme are proposed.EZC-DCT compression algorithm has the advantages of both JPEG and JPEG2000,so it can build a multi-resolution terrain model quickly,which is similar to the LOD level detail model,and can build terrain models with different resolution levels in different range range dynamically according to the human eye characteristics.At the same time,combine the fast EZC-DCT algorithm,“step + transform” multi-resolution terrain model construction strategy and GPU parallel acceleration scheme,so as to achieve quickly real-time compression and decompression of terrain data and multi-resolution terrain model construction.The main work is as follows:1.Further understanding of the research on image compression technology and the EZC-DCT algorithm.While The fast algorithm of DCT has been studied,and the basic process of each fast algorithm is mastered.On this basis,a fast EZC-DCT transformation algorithm based on integer is proposed.2.The algorithm and current research results of terrain real-time rendering are deeply studied.At the same time,the LOD algorithm and the construction strategy of multiresolution terrain model based wavelet LOD have been studied.On this basis,a strategy of "step + transform" terrain model is proposed,which can further reduce the number of transformation and improve the real-time rendering frame rate.3.Based on the original algorithm,a parallel optimization scheme of GPU is proposed.The original EZC-DCT method performs serial processing and transformation to each block by dividing the original image into blocks,so the operation efficiency is low.Compared with CPU,the processing speed of GPU is faster,so parallel optimization of EZC-DCT algorithm will greatly reduce the load of computer,and it can effectively improve the real-time rendering frame rate.