论文部分内容阅读
运用模式识别技术,提出了分数阶微分自适应去噪算法,它基于AA模型(Aubert-Aujol模型)运用分数阶微分理论对图像进行建模,建立了分数阶微分模型,模型参数分数阶阶数u和图像各点对应正则化参数λ根据区域特征进行自适应选择。数值实验结果表明,与传统去噪方法比较,应用分数阶微分自适应去噪算法后,其衡量去噪效果的定量指标-峰值信噪比(PSNR)和边缘保持指数(EPI)得到有效的改善,获取的数值效果较好。
Using pattern recognition technology, a fractional differential adaptive denoising algorithm is proposed. It uses the fractional order differential theory to model the image based on the AA model (Aubert-Aujol model), and establishes the fractional differential model. The model parameters fractional order u and regularization corresponding to each point of the image λ according to the regional characteristics of adaptive selection. Numerical experiments show that compared with the traditional denoising method, the PSNR and EPI, which are quantitative indicators of denoising effect, are effectively improved after applying the fractional-order differential adaptive denoising algorithm. , Get better numerical results.