考虑梯度边缘信息的分形图像纹理细节增强算法研究  

Research on Fractal Image Texture Detail Enhancement Algorithm Considering Gradient Edge Information

作　　者：胡敏 张海民 程菲 HU Min;ZHANG Hai-min;CHENG Fei(Xuancheng Vocational And Technical College Department of Information Engineering,Xuancheng 242000,China;Anhui Institute of Information Technology School of Computer and Software Engineering,Wuhu 241000,China)

机构地区：[1]宣城职业技术学院,安徽宣城242000 [2]安徽信息工程学院

出　　处：《通化师范学院学报》2022年第10期6-12,共7页Journal of Tonghua Normal University

基　　金：安徽省自然科学基金资助项目(2008085MF201);安徽省教育厅自然科学重点研究项目(KJ2021A1206)。

摘　　要：为降低噪声干扰,增强图像纹理细节,提高图像质量,提出考虑梯度边缘信息的分形图像纹理细节增强算法.依据图像已知点的灰度值推测新增点的灰度值,将新的像素点插入在图像像素点间,通过FBR理论对图像进行分形插值处理;采用图像锐化技术改进消噪掩模法,获取分形图像梯度边缘信息;构建梯度约束与数值约束的增强模型,控制图像噪声的同时,增强分形图像的边缘纹理细节.实验结果表明:信噪比越大,该算法分形插值处理后图像分辨率越高;该算法图像纹理细节增强后的图像噪声含量低,图像纹理与结构上相似度高,图像纹理细节呈现效果好;去噪能力较强,可有效增强图像边缘与细微纹理细节.In order to reduce noise interference,enhance image texture details and improve image quality,a fractal image texture detail enhancement algorithm considering gradient edge information is proposed.According to the gray value of the known points in the image,the gray value of the new points is inferred,the new pixels are inserted between the image pixels,and the image is fractal interpolated by FBR theory.The image sharpening technology is used to improve the denoising mask method to obtain the gradient edge information of fractal image.The enhancement model of gradient constraint and numerical constraint is constructed to control the image noise and enhance the edge texture details of fractal image.The experimental results show that the higher the signal-to-noise ratio,the higher the image resolution after fractal interpolation.The algorithm has the advantages of low noise content,high similarity between image texture and structure,and good presentation effect of image texture details.It has strong denoising ability and can effectively enhance image edges and fine texture details.

关 键 词：梯度边缘信息 分形图像 纹理细节增强 噪声干扰 分形插值 FBR理论 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]