摘要:This paper presents a new method for fast fractal image coding. This method speeds up the coding process by adaptive partition of image, moment classification of domain pool, orthogonal decomposition of domain blocks and range blocks, as well as dynamic search by pixel after pixel. The experimental results show that the method presented in this paper performs well in speeding up the coding process, especially for those images that have much more details in them.
摘要:The method of improving the PSNR in fractal compression is studied.Based on the magnification effect of fractal code for image, the algorithm extracts the high frequency information in the zooming out image by wavelet transform, then merges the information into the decoded image by wavelet transform. A modification method of Lipschitz exponent of wavelet coefficients local module maximum is also investigated in order to get rid of the blocking effect in the result image. This kind of method is used to improve the PSNR of result image in low bits rate image compression.
摘要:Based on mathematical morphology, this paper analyses texture feature of B-scan ultrasonic images of liver and presents an algorithm for rotation invariant texture feature extraction. Compared with other morphology-based algorithms, experimental results show that our texture feature has high classification accuracy.
摘要:In accordance with much too slow speed drawback on generation of virtual scenes based on particle systems, this paper first describes general principle of particle systems; then analyzes property of rain and snow; finally presents a real time simulating algorithm of rain and snow descending based on particle systems.
摘要:Finite\|length input signal will introduce extension or distortion in wavelet transform. Analyzing the impact of symmetric extension on analysis/synthesis system, this paper presents optimal symmetric extension of finite\|length input signal and delay of filters in subband coding which produces perfect reconstruction and overall zero\|delay.
摘要:This paper firstly made a brief introduction of the technology for information fusion, and also summarized the previous and resent worldwide studies on the fusion of multisensor satellite remote sensing data. Then, aimed at geological application, the paper exhibited the utility of wavelet transform for data fusion purpose illustrated by an example of Landsat-TM and SPOT-PAN images. On the one hand, the authors performed a discussion on the relationship between wavelet coefficients and the fused results, and set up a fusion model called“CMWD”; On the other hand, the paper raised a comparison of the fused images which were merged by wavelet transform, PCA and IHS methods, in spatial quality improving and spectral information preserving respectively. Finally, based on Multi-Agent theory, the paper presented a framework for building up an information fusion system of satellite data.
摘要:Based on the topological and adjacent properties of graph, shape configuration of nodes generated by adaptive raster to block procedure is analyzed in detail and a new approach for recognition of crossing area in scanned engineering drawings is proposed in this paper.
摘要:This paper presents a visualization method of 3D finite element analysis. The high stress zone is filled with translucent material and the low stress zone is filled with transparent material.In this way, the 3D image of high stress zone, both the shape on surface and the depth of the zone, can be observed directly.
摘要:Ray casting is a classic algorithm of Volume Rendering. This method has the advantages of clear structure and easy to be accomplished. But simplify ray casting algorithm exists some weakness: low resampling efficiency and low image accuracy. This paper utilizes the coherence of 3D data filed and resampling with unequal step to improve the image quality of traditional ray\|casting method and reduce the rendering time.
摘要:The triangulated irregular network(TIN) is a very useful model for data compression and 2 5D analysis.Unfortunately,published algorithms of building multiresolution TIN can't meet the demand of adjusting precision momentarily.This paper presents a set of algorithms for building multiresolution TIN.It covers algorithms of constructing fundamental TIN,of adding points to the current TIN and of removing points from the current TIN.All algorithms were encoded with MSC.Two versions of executable program were provided.One for creating precision determined TIN,the other for adjusting the precision when running.Test result proves the algorithms are correct and the programs are stable and reliable.
摘要:In this paper, discussing the method in auto extracting wave from seismic record, a robust horizontal correcting and a horizontal line deleting algorithms are presented. The authors also produced a variational Hough transform used for wave correcting, and concluded the wave model for wave extracting. After that,the rules in intellective wave searching method are presented.
摘要:The effect of DSA(Digital Subtraction Angiography) images acquired with general subtraction algorithm is not satisfied, the merit of DSA images cannot be expressed and used extensively. According to the principle of formation of X ray image, this paper described an improved subtraction algorithm called the logarithm non linear grey mapping. In comparison with general subtraction algorithm, this improved subtraction algorithm finally acquired a satisfactory sequence of DSA vascular images with high contrast and high definition.
摘要:“Rainbow Range Finder”is a method for rapidly acquiring 3D information based on spectral analysis. It uses a special light with continuous spectrum to project across the objects, and its image will present the regular change of colors. One color forms a line in the color image and responds to a light plane structured by a wave band in the spectrum. When all the light planes are calibrated and the camera model is known, we can calculate 3D coordinates of all image points in the scene. This paper mainly discusses the light plane calibration and color classification techniques for implementation of these methods.
摘要:Based on the intensity model of natural landscapes with steam environment in the HVS display method, the perspective cloud discrete model is formed to determine the transformation relations between the parameters of R、G、B(red,green,blue) and the intensity in the RGB display method. The highly speed model of perspective cloud is presented. This method has such characteristics as high speed, good effect for display.
摘要:In this paper, wavelet analysis is utilized to analyze and process the aero magnetic anomality of Sheng Li Oil Field. The wavelet property of aero magnetic of the oil field is studied, and the relationship between this property and the disrtibution of the underground oil is found. The result indicates that the distribution of underground oil location is the same as high frequency part derived from the aero magnetic anomaly signal.Furthermore,high frequency part is helpful to understand the controling factor of oil savings. Using this method,the authors predict by the other high frequency parts,and the results are acknowledged by the oil geoprahy experts working in local area.
摘要:In this paper,based on the method of block coding fractal image compression proposed by Jacquin, a new fractal color image compression method based on Vector Distortion Measure and Adaptive Quadtree Partiton is presented.This method fully exploits the correlation between the three color components of 24 bits true color image in RGB color space,experiments show that both compression ratio and reconstructed image quality are much better than that of separate fractal coding in RGB color images.
摘要:In this paper, we pay attention to the application of image compression with biorthogonal wavelet transform. We provide an algorithm which is suited for fast progressive and lossless compression.