基于小波双立方插值的Bessel拟合图像复原算法

对模糊图像复原的目的在于对图像进行去模糊、去噪及建立在特征值基础上的矢量化处理，最大限度恢复原图像。本文提出了一种基于小波双立方配比插值的贝塞尔拟合图像边缘特征复原的新算法，在初始图像降质处理基础上通过小波双立方配比插值（Wavelet Bi-cubic）、双三次插值法得到融合样本图像，并在Canny边缘检测基础上提出对拟合边缘特征点的贝塞尔曲线拟合，最后将图像特征轮廓分别进行Delaunay剖分算法的矢量化处理，保留原始灰度色彩复原图像。实验对这三种方法的轮廓复原图像与原始清晰图像进行对比，由具体参数（PSNR、SSIM、FSIM以及运算时间TOC等）说明验证。实验结果表明，经由小波双立方配比插值与贝塞尔拟合得到的轮廓图像复原比双三次插值与贝塞尔拟合得到的轮廓图像复原效果要好，且明显缩短运算时间。

Bessel fitting image restoration algorithm based on Wavelet Bicubic Interpolation

The key to restoration of blurred image is to remove blur and noise from the image and to vectorize the image based on eigenvalues, so as to restore the image to the maximum extent. In this paper, a new edge feature restoration algorithm relating to Bessel fitting on Wavelet Bicubic ratio interpolation is proposed. On the basis of the initial image degradation processing, the fusion sample image is obtained by wavelet Bi cubic interpolation and bicubic interpolation. On the basis of Canny edge detection, the Bezier curve fitting of the fitting edge feature points is proposed. Finally, the image feature contour is vectorized by Delaunay segmentation algorithm respectively to retain the original gray color restored image. The experimental results show that the three methods are compared with the original clear image, and the parameters (PSNR, SSIM,FSIM and TOC) are used to verify the results. The experimental results show that the contour image restoration by wavelet bicube ratio interpolation and Bessel fitting is better than that by bicubic interpolation and Bessel fitting, and the operation time is obviously shortened.