ID 原文 译文
10304 同时,为了克服CLMS算法停滞等待的弊端,采用了瞬时转移结构; ‭At the same time, in order to overcome the disadvantages of CLMS algorithm stagnation waiting, the instantaneous transfer structure;
10305 ‭另外,在参数的迭代公式中使用sign函数进行优化以降低运算量。 ‭In addition, the parameters used in the iterative formula of sign function is optimized to reduce the computational complexity.
10306 仿真结果证明该算法与CLMS、VS-CLMS相比,在不同的仿真环境中均能表现出良好的均方特性和跟踪特性。 ‭Simulation results prove that the algorithm is compared with CLMS, v - CLMS, in the different simulation environment can show the good characteristics of mean square and tracking.
10307 为了提高偏微分方程放大算法对纹理细节的放大效果,利用改进的复扩散模型耦合均值滤波器,提出了一种图像放大算法。 In order to improve the algorithm of texture details of the partial differential equation of amplification of amplification effect, using the improved complex diffusion model coupling the mean filter, puts forward an algorithm of image magnification.
10308 改进的非线性复扩散模型能够很好的定位图像边缘,通过冲激滤波器对边缘进行锐化,同时耦合非局部均值滤波器,保持图像内部的自相似特性,利用非局部信息重建高分辨率图像,提高小边缘、细节放大效果。 Improved nonlinear complex diffusion model can better location image edge, by impulse filter to sharpen edges, coupling nonlocal average filter at the same time, keep the image within the self-similar characteristic, high resolution image reconstruction using nonlocal information, improve the effect of small edges and details amplification.
10309 算法结合非局部信息和局部信息对图像进行放大,增强纹理细节,使图像更加自然,同时减弱对边缘的过度增强,具有较好的放大效果,仿真实验验证了算法的优良性能。 Algorithm combining the local information and local information of image to enlarge, strengthen the texture details, make the image more natural, less on the edge of the excessive increase at the same time, has good amplification effect, the simulation results demonstrate the good performance of the algorithm.
10310 根据贝叶斯分类准则提出了一种改进的基于局部与全局信息的水平集图像分割模型。 According to the bayesian classification criterion is proposed an improved level set image segmentation based on local and global information model.
10311 首先,利用图像的局部信息建立了局部能量项,引导目标附近的演化曲线停在目标边缘上; ‭First of all, the local information of image is used to establish the local energy item, lead to the edge of the curve evolution stops at target near the target;
10312 然后,利用图像的全局信息建立了全局能量项,加速远离目标边缘处演化曲线的演化; Then, the global energy was established based on global information of the image, accelerate away from the evolution of the target edge curve evolution;
10313 最后,提出了一种联合局部能量项和全局能量项的统一的水平集模型架构,提高了分割效率和分割灰度不均匀图像的能力。 Finally, this paper proposes a joint local energy and the unification of the global energy item level set model structure, improve the efficiency of segmentation and uneven gray scale image segmentation.