最新刊期
卷 21 ， 期 4 ， 2016
- 摘要：Local and international survey research on the convex energy functional regularization model(EFRM) are rare. To obtain comprehensive understanding of potential research, research development is surveyed in the field. Based on numerous references, we summarize and compare convex EFRM from four aspects, namely, original, composition, treatment, and development. First, given the inverse problem, obtaining feasible solutions is impossible. The effective method of solving the problem is to establish EFRM. Second, after establishing the energy functional model, the application conditions of the fitting and regularization terms are analyzed. Five types of point spread functions that can make image blur are provided. The significance and weight determination method are presented. Third, the fitting and regularization terms of the energy functional model are sorted by entire and separate treatments, and the computational algorithms of the regularization model are analyzed in space, transform, and hybrid domains. The advantages and disadvantages of the models and algorithms are determined. Finally, the development trends and existence problems of EFRM of image restoration are highlighted. Generally, directly solving primal convex EFRM is impossible; however, with transformation model, dual model, and primal-dual model of the primal model and by taking advantage of numerical analysis, matrix theory, and optimization theory, designing effective and rapid algorithms to completely and separately solve the transformation model becomes possible. Meaningful theories and applicable results of EFRM of image restoration have been obtained. However, with the continuing appearance of large-scale data processing problems, several theoretical problems, such as establishment of accurate mathematical models, design of effective and fast solving algorithms, and analysis of algorithm convergence properties, require further research.关键词：energy functional regularization model;image restoration;optimization algorithm design;large scale data processing3003|1622|2更新时间：2024-05-07
- 摘要：Image sparsity cannot be fully exploited through either whole or fixed block sampling during image compressive sensing(CS). An insurmountable discrepancy is observed between sampling rate and image reconstruction quality. This study proposes an adaptive block compressive sensing(ABCS) algorithm based on the changes of image texture. The algorithm combines JPEG quantization methods and reduces the sampling rate to increase the compression ratio without reducing the quality of the premise of image reconstruction. The proposed ABCS algorithm requires that the image be roughly divided into blocks for starters. Then, the texture complexity of each block is calculated and analyzed to select the optimal sampling rate in the CS process. Whether a block should be further divided depends on whether its texture complexity is higher or lower than the corresponding threshold. When the texture complexity is lower than the corresponding threshold, the optimal sampling rate should be selected to apply compressive sampling. When the texture complexity is higher than the corresponding threshold, the block should be further divided into even smaller blocks. The process is repeated until the block size is 16×16. After the division, if the texture complexity is still higher than the maximum corresponding threshold, then JPEG quantization coding should be applied. The new compression algorithms are generated by combining the ABCS and typical CS reconstruction algorithms. Experiments were conducted to compare the new algorithms and the original algorithms. Under the condition of similar sampling rate, the quality of the reconstructed image is evidently improved, particularly at a low sampling rate. For example, the peak signal-to-noise ratio of the reconstructed image is approximately 30 dB at 20% sampling rate. Simulation results show that adaptive block sampling makes full use of the sparsity of the image and improves the efficiency of compressed sensing. The blocks using JPEG encoding avoid poor-quality reconstruction of the complex texture area and reduce the reconstruction time.关键词：image compression;compressive sensing;sampling rate;sparse representation3652|388|2更新时间：2024-05-07
- 摘要：A class of cubic trigonometric pseudo B-spline functions combining multiple characteristics is proposed to address the deficiency of normal polynomial B-spline curves in shape control, the accurate representation of several engineering curves, and the construction of interpolation curves. A class of cubic trigonometric pseudo B-spline basis functions with two parameters is constructed. The parametric pseudo B-spline curve is defined based on the proposed basis functions, and the characteristics and fairness of the curve are discussed. Lastly, the corresponding algebraic pseudo B-spline is studied, and a method of determining the optimal algebraic pseudo B-spline is provided. The parametric pseudo B-spline curves can not only be C continuous, but can also automatically interpolate the given data points without solving the system equations. When the data points are fixed, the shape of the interpolation curves can be adjusted by the two parameters. Under appropriate conditions, the parametric pseudo B-spline curves can be precisely expressed as several engineering curves, such as circular arc, elliptical arc, and star line. The corresponding algebraic pseudo B-spline has properties similar to those of the parametric pseudo B-spline curve. The optimal algebraic pseudo B-spline can obtain satisfactory interpolation effects. The proposed pseudo B-spline simultaneously solves the deficiency of normal polynomial B-spline in shape control, the accurate representation of several engineering curves, and the construction of interpolation curves. Thus, it is a practical method for curve modeling.关键词：B-spline;trigonometric spline;interpolation spline;shape adjustment2279|355|2更新时间：2024-05-07
- 摘要：Accurate and precise vessel segmentation is the key technology of medical image three-dimensional(3D) reconstruction and visualization. However, the existing methods of vascular segmentation still have shortcomings. According to the correlation between adjacent layers of computed tomography(CT) images and multiple directions of vascular distribution, this study proposes a new 3D vessel segmentation algorithm based on the Stein-Weiss analytic function. The algorithm can delineate more precise and clearer blood vessels, particularly in the case of noise and other effects of broken vessels. First, the contrast of vascular point and background was increased by pretreatment of image through enhancement and adjustment to window width and position. Then, we combined the Stein-Weiss function with the gradient operator. Each voxel of the CT volume data was expressed as a Stein-Weiss function. We also used the gray value of the six neighborhoods of the voxel as coefficients of the Stein-Weiss functions. Then, we determined the gradient values of the Stein-Weiss function in the three directions of , , and . When the values were larger than the threshold, the voxel was considered a point of the vascular edge. Finally, according to the CT slices extracted from the edge of the vessels, we reconstructed the 3D blood vessels. In the experimental part, we used our proposed algorithm, the octonion analytic function, and the region-growing algorithm to segment blood vessels. Given that the region-growing algorithm cannot continue to grow in the region of the breakpoint of the vascular region, the first two methods can delineate more abundant blood vessel branches. As the Stein-Weiss analytic function has better analytic properties than the octonion analytic function, the segmentation of blood vessels using our proposed algorithm is clearer and has lesser noise than the segmentation using the octonion analytic function. The experimental results of liver blood vessel segmentation and 3D reconstruction show that the sensitivity and specificity of the proposed algorithm are relatively high compared with that of the region-growing algorithm and the octonion analytic segmentation algorithm. In particular, the algorithm has evident advantages in the denoising of vessel segmentation and can rapidly and efficiently segment clearer and more refined vessels. In this study, a new algorithm for vessel segmentation is proposed. We used the Stein-Weiss function of high-dimensional mathematical tools to extract vessel edge. The proposed algorithm can consider multiple characteristics during the calculation process. Compared with the results of a small number of characteristics, the proposed algorithm has evident advantages and wide application prospects.Because the nature of Stein-Weiss parsing can be suitable for arbitrary dimension, we can use the Stein-Weiss analytic function to identify the image edge of 2D or higher dimension.关键词：Stein-Weiss analytic function;pretreatment;octonion;vessel segmentation;three dimension reconstruction2784|478|2更新时间：2024-05-07
- 摘要：Intravascular photoacoustic(IVPA) imaging combines photoacoustic imaging and medical endoscopic technique. It provides reliable information on cardiac vessels and atherosclerotic plaques for the detection of vulnerable plaques and interventional treatment of cardiac diseases. A method of reconstructing cross-sectional gray-scale images based on the time-reversal algorithm is proposed for the IVPA imaging system equipped with a single ultrasonic detector with circular scanning. The back-propagation of photoacoustic waves generated by the vascular tissues is modeled and numerically simulated to obtain 2D images of photoacoustic pressure distribution. The photoacoustic signals collected by the ultrasonic detector are interpolated to solve the problem of incomplete photoacoustic data caused by sparse measuring positions and limited-view measurement. The results of computer-simulated experiments suggest that the structural similarity of the reconstructed images is at least 65% higher than that of images obtained with the filtered back-projection algorithm. The proposed method can effectively improve the quality of the reconstructed image for IVPA imaging and can serve as a useful reference for subsequent optimization of image reconstruction algorithms.关键词：intravascular photoacoustic(IVPA) imaging;image reconstruction;time-reversal(TR);filtered back-projection(FBP);structural similarity3836|484|3更新时间：2024-05-07
- 摘要：Current research on the 490 nm diffuse attenuation coefficient are mostly about the building of inversion model at 490 nm band, but has no connection with airborne LIDAR hydrographic ability. This study attempts to determine the hydrographic parameters (532) and , which could be used to estimate the ability of the airborne double-color hydrographic LIDAR system and the important parameters for the flying scheme. First, the importance of the diffuse attenuation coefficient (532) and as the hydrographic parameters for the airborne double-color hydrographic LIDAR system is analyzed. Then, according to the in situ measured data in the Yellow Sea and East China Sea coastal areas and the existing diffuse attenuation coefficient retrieval algorithms modified in this study, the linear correlation of (532) and (490) and the exponential correlation of and (532) are derived. Based on the Moderate Resolution Imaging Spectroradiometer(MODIS) (490) products and the aforementioned functions, the hydrographic parameters (532) and are obtained. The hydrographic parameters derived by the method employed in this study can be used availably to estimate hydrographic ability, whose accuracy depends on the accuracy, distribution, and quantity of measured optical data and the accuracy of MODIS (490) products.关键词：MODIS;diffuse attenuation coefficient;band;Secchi disk depth;color scale2668|349|2更新时间：2024-05-07
- 摘要：The data collected by synthetic aperture radar(SAR) are sparse and incomplete because of the imaging method and geometry angle, which result in difficulties in 3D SAR data surface reconstruction. For downward-looking SAR data, this paper presents a reconstruction method that can repair sparse and incomplete data automatically. First, the Otsu method is introduced for the preprocessing of 3D SAR data. Second, the Chan-Vese model of 2D image segmentation method is applied to the surface reconstruction of the downward-looking SAR data. The distance and inner product functions are employed as the initial surface and contour indicator function. Compared with the isosurface extraction method, the proposed method can repair the holes automatically during the reconstruction procedure. The reconstructed model surface is smoother and can reflect the characteristics of the original object. The proposed method can be applied to the modeling of sparse and incomplete data.关键词：downward-looking SAR(synthetic aperture radar) data;hole-filling;3D reconstruction;Chan-Vese model2474|356|0更新时间：2024-05-07
- 摘要：Fog-degraded images seriously affect the normal operation of the information systems of the military, traffic, safety monitoring, and other fields. Hence, research on image dehazing has become important and practical. Most existing single image dehazing approaches focus on investigating haze-relevant color features. However, different color priors have their respective limitations. To improve the adaptability of the image dehazing algorithm, we propose a single image dehazing algorithm by feature learning. First, we extract multi-scale textural and structural features from a hazy image by using a sparse auto encoder; the haze-relevant color features are obtained as well. Second, we model the scene depth with the extracted textural, structural, and color features by using multi-layer neural networks. By training samples with the neural network model, the relationship between scene depth and the features is determined, and the depth map of the hazy image is restored. Finally, the haze-free image with the depth map is restored according to the atmospheric scattering model. Compared with the results of mainstream algorithms, those of the proposed algorithm are more natural and have more details. Qualitative evaluation of the similarity between our results and the ground truth haze-free images indicates that our results have high similarity to the ground truth situation, reaching 99.9%. Experiments show that the proposed algorithm can effectively restore the depth map of a hazy image and produce a high-quality haze-free image with good adaptability.关键词：image dehazing;feature learning;deep learning;neural networks;sparse coding6841|561|6更新时间：2024-05-07
- 摘要：Edit propagation is an important operation in image processing. Previous methods require that the samples selected interactively should have a sufficiently large number of colors. In this study, we propose a new robust edit propagation technique by efficient resampling, which may reduce the requirement of interactive operation. First, the image is segmented into superpixels and the center pixel of each superpixel is considered a new sample. Following the rule that each sample obeys one kind of editing mode, we calculate the edited color of each new sample by the similarity between the sample and editing modes and the similarity between those samples and its neighbors. The sample whose similarity to several editing modes is evidently high directly belongs to the corresponding editing mode. Otherwise, the editing mode of the current sample is determined by the neighboring samples with high similarities to the current sample. Then, the edited colors of all samples are propagated into the entire image based on the policy that similar colors are still similar after editing. Based on the similarity of the pixel and the samples, we select the number of samples adaptively for each pixel to reduce the interference of those samples with low similarity. Compared with previous methods, our solution is less sensitive to the input strokes. With the same simple strokes, our results exhibit better vision and details. The method can effectively simplify the user interaction by replacing the sparse input samples with many colors resamples. Moreover, the experiments and comparisons show that the method can preserve the edited colors and its details.关键词：image editing;superpixel segmentation;sparse interactive;edit propagation;detail-preserving4021|424|1更新时间：2024-05-07
- 摘要：Image interpolation plays a vital role in image processing. A new image interpolation algorithm based on a rational fractal function is proposed to improve the quality of texture image magnification. This method is combined with a previous rational function interpolation algorithm. For input image preprocessing, a median filter and histogram equalization are utilized. The texture and smooth areas in the image are classified through the blanket method and the multi-scale fractal characteristic value of the image. Finally, a rational fractal interpolation function is employed for the texture region, and a rational interpolation function is adopted for the smooth area. An optimization technique is then utilized to further modify the interpolation model, which is proven to be effective. A rational fractal interpolation algorithm is proposed in this article. For common images, the quality of interpolation approximates that of NEDI and NARM. For texture images, the proposed method is highly competitive not only in PSNR and SSIM but also in visual effect. This article presents a novel image interpolation method based on a rational fractal function. Experimental results demonstrate that the proposed method exhibits competitive performance, especially in terms of image details and texture features.关键词：fractal;image interpolation;multi-scale analysis;fractal dimension;rational fractal;gradient4862|324|3更新时间：2024-05-07
- 摘要：Compressed sensing signal reconstruction involves finding the sparse solution of an underdetermined system of linear equations. This problem has three common solutions:minimizing the l-norm, l-norm, and l-norm. Minimizing the l-norm can solve this problem, but the method belongs to NP complete. Meanwhile, the l-norm has no solution in theory, and the l-norm is not a convex function. Under this background, this paper presents a recovery method of compressed sensing signals based on regularized smooth convex optimization. To obtain the global optimal solution, a convex function in the entire space is first designed as the objective function of optimization to fit the l-norm of a signal. Second, the optimization problem with variables is transformed into optimization problems with one variable. Finally, a fast iterative shrinkage-thresholding algorithm is proposed to find the solution, with the convergence speed being that of quadratic convergence. Experimental results show that the method is robust and fast compared with three other types of algorithms. Specifically, the reconstruction time of the proposed method is approximately 50% of that of the other algorithms when the signal has more than 150 dimensions. In the real data experiments, the F-norm of the reconstructed signal and the original signal difference with the proposed method is approximately 70% of the other algorithms. The proposed algorithm with a high compression ratio and a good restoration effect is well-suited for processing large data. It meets the requirements for application in information retrieval, dictionary learning, image compression, and so on.关键词：signal processing;compressed sensing;convex function;fast iterative shrinkage-thresholding;sparse representation4099|301|3更新时间：2024-05-07
- 摘要：Given that the traditional quantum genetic algorithm cannot make full use of immature individuals in a population, we proposed and used the quantum genetic algorithm based on interactive update mode(IUMQGA) in solving geometric constraints. The constraint equations of the geometric constraint problem can be transformed into the optimization model; therefore, constraint-solving problems can be transformed into the optimization problem. A quantum genetic algorithm, which combines genetic algorithm and quantum theory, using the double strand quantum chromosome structure, is employed to achieve the crossover operation of the genetic algorithm using quantum gate transformation. According to different situations, different interactive update strategies are adopted. The term "interactive" refers to the process of information exchange between two individuals. The process is used to generate new individuals. The IUMQGA uses the interactive update mode to achieve the crossover operation of the genetic algorithm using the quantum gate transformation. The process not only increases information exchange between two individual but also makes full use of the information of immature individuals and improves the converge speed of the algorithm. The comparison of nonlinear equations and geometric constraints with other methods shows that IUMQGA for solving geometric constraint problems has better accuracy and solving rate. The IUMQGA algorithm is more stable than the QGA algorithm in the case of the double circle tangent problem. The experiments reveal that the error of the optimal value of the variables and the corresponding accuracy is below 1E-2. The IUMQGA can be used to solve the geometric constraint problem.关键词：computer aided design;geometric constraint solving;quantum genetic algorithm;quantum chromosome;interactive update mode4086|373|2更新时间：2024-05-07
- 摘要：The root-finding problem has a wide range of application in computer-aided design and computer graphics. This study attempts to improve conventional clipping. For a multiple root case, good computational stability is maintained and a higher convergence rate is achieved. This study presents a new clipping method based on space. The proposed method utilizes the good computational stability of Bernstein basis functions, provides a simple method to evaluate the existence of multiple roots, and converts a multiple root case into a simple root case. Compared with conventional cubic clipping methods based on and spaces, the proposed method achieves a better approximation effect. The proposed method can also achieve the convergence rate of 5 for a multiple root case of multiplicity m, which is better than the convergence rates 4/ and 5/ of previous cubic clipping methods. The computational complexity of different cubic clipping methods are close. The new method can evaluate the existence of multiple roots and can achieve a higher convergence rate and a better approximation effect for a multiple root case.关键词：computer graphics;root-finding of polynomials;cubic clipping method;robustness;R;space;approximation order;multiple root4254|305|0更新时间：2024-05-07
- 摘要：In several practical interpolation problems, the integral values of several subintervals are known, whereas the usual function values at the knots are not given. In this article, we propose an integro quadratic spline interpolation method from the integral values of successive subintervals. First, we use the linear combination of quadratic B-spline functions to satisfy the given integral values and two additional boundary conditions. This problem is equivalent to solving a system of +2 linear equations with a three-band coefficient matrix. Second, we use operator theory to conduct an error analysis and obtain the super convergence order in approximating the function values at the knots. Lastly, we deal with the integro interpolation problem even without any boundary condition based on the linear combination of integral values to approximate the boundary function values. The proposed and modified methods are tested by functions with low frequency in Example 1; the approximation behavior are satisfactory. Meanwhile, the proposed method is tested by functions with high frequency in Example 2; the numerical convergence order is consistent with the theoretical value. The proposed method is easier to implement than other existing methods. The super convergence in approximating function values at the knots is also verified. The proposed method of function reconstruction from the integral values of successive subintervals is universally applicable.关键词：interpolation;quadratic B-splines;integral values;boundary conditions;error analysis4243|377|1更新时间：2024-05-07
- 摘要：Line of curvature plays an important role in differential geometry. It is a useful tool in surface analysis. A developable surface is simple and frequently used in surface modeling. Many studies have investigated how to find or compute the line of curvature from a given surface; its anti-problem is rare. This paper presents an approach of designing a developable surface by interpolating the line of curvature and applying it to surface modeling. The Frenet frame is employed to express the ruling of the ruled surface. According to the necessary and sufficient conditions of line of curvature and developable surface, the equations that satisfy the ruling can be obtained. The developable surface can be adjusted flexibly by introducing control functions. The paper presents the concrete expression of a developable surface. The necessary and sufficient conditions are analyzed by classifying the developable surface. The method is validated by two representative examples. Results show that the method is suitable not only for general parametric curves but also for piecewise parametric curves. The concrete expression of a developable surface is constructed by interpolating the line of curvature through a constructive method. The proposed method is efficient and convenient, as verified by several representative examples.关键词：developable surface;ruled surface;frenet frame;line of curvature3940|224|2更新时间：2024-05-07
- Postal code：100190
- Tel：010-58887035/58887030/58887418 Email：jig@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备05080539号-4
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0