摘要
针对图像放大时分辨率会降低,图像的感知质量变差等问题,并考虑到计算的复杂度,提出了一种差分投影快速定向图像插值算法。该方法通过对低分辨率图像的梯度扩散确定高分辨图像的丢失像素,由特殊的去噪方法对该算法进行优化,并与最近邻插值法与双线性插值算法等相比较。实验结果表明,该方法具有较高的分辨率和较低的计算复杂度,优于其他方法。
A fast differential projection image interpolation duced image resolution and poorer quality of perception in image algorithm is proposed to address the problems of re- magnification with complexity of calculation in con- sideration. The loss of the high resolution image pixels is determined by low resolution image gradient diffusion and the algorithm is optimized by a special denoising method and compared with the nearest neighbor interpolation method and bilinear interpolation algorithm. The experimental results show that the method has higher resolution and lower computational complexity than other methods.
出处
《电子科技》
2015年第6期28-32,共5页
Electronic Science and Technology
基金
沪江基金资助项目(C14002)
关键词
图像放大
差分投影
梯度扩散
分辨率
去噪
image magnification
difference projection
gradient diffusion
resolution
denoising
