双线性插值
图像缩放
放大倍数
脱模
插值(计算机图形学)
图像纹理
算法
阶梯插值
数学
图像(数学)
图像处理
特征检测(计算机视觉)
计算机视觉
人工智能
计算机科学
二值图像
多元插值
作者
Lifang Xiao,Dacong Jin,Yong‐Xian Cheng,Mingyang Luo
标识
DOI:10.1109/icrcv55858.2022.9953257
摘要
Image magnification is an image processing technique that obtains high-resolution images from low-resolution images. This paper is proposed a combined image enlargement algorithm of fractional differential and bilinear interpolation, which makes up for the defect that the texture of the image cannot be saved compared with the previous integer order. First, the original image is reduced to one-half, and then the image is magnified by two times by bilinear interpolation, and then the image magnification algorithm is constructed by using the fractional order differential operator to enlarge the image, and the image magnification algorithm of the combination of the fractional partial differential operator and the bilinear interpolation is constructed. Experimental results show that the combined image magnification algorithm can maintain the basic characteristics of the image while better maintaining the texture features of the image detail edges.
科研通智能强力驱动
Strongly Powered by AbleSci AI