小波与自蛇模型在放大图像清晰化中的应用

针对传统插值放大图像出现的边缘模糊与锯齿化问题，结合小波具有多分辨率分析和局部化时频域特性，提出了一种基于小波插值与改进自蛇模型相结合的放大图像清晰化方法。该方法对无噪声图像采用小波插值对图像进行放大，并用改进自蛇模型对放大后的图像进行边缘修正，而对于噪声图像则采用改进自蛇模型对其进行清晰化处理，通过小波插值进行放大。实验结果显示，采用该方法与传统放大图像清晰方法相比，图像的边缘轮廓清晰度和细节部分的辨识度更精确，同时能够有效提高放大图像的峰值信噪比。

ZHANG Kaige,et al. Wavelet and self-snake model applied to sharpen amplified image

In view of blurred and jagged edges in the image magnified by traditional interpolation, combining with multi-resolution analysis and time-frequency localization characterized by wavelet, a sharpening method based on wavelet-inter-polation and improved self-snake model is proposed. This method zooms in the image with the combination of wavelet-interpolation algorithm and uses improved self-snake model to correct the edge for magnified image, while in terms of noise image, an improved self-snake model is used to sharpen the image, and then enlarge the processed image by wavelet-interpolation. Experimental results show that this method has a more accurate sharpness of the edge profiles and a higher recognition of the detail section compared with the traditional sharpening algorithms in enlarged images, at the same time, it can effectively improve the PSNR of the enlarged image.