提出了一种以图像区域的一致性为指导的自适应选择插值公式(简称ADA)的图像插值算法。首先将图像分割成不同区域,然后结合近邻法和众数法确定待插值点所属区域,将待插值点分类处理:对于区域内的点,采用线性插值;对于区域间的过渡点,设计非线性插值公式,分配较大的权值给与待插值点属于同一区域的邻域像素,分配较小的权值给其它邻域像素。实验表明,ADA图像插值算法得到的插值图像有较好的主观图像质量,其峰值信噪比(PSNR)比传统方法平均提高了3.11~4.39dB。
An area-directed adaptive image interpolation algorithm(ADA) was proposed. The area that an interpolated pixel should belong to was decided by the approach combining the method of the nearest neighbor and the statistical mode. For transition pixels between different areas,nonlinear interpolation formulas were designed. Bigger weights were assigned to neighboring pixels that have larger contributions to calculate the interpolated point value. The experimental results show that ADA can obtain interpolated image with good subjective quality,and the average PSNR is improved 3.11 to 4.39 dB compared with that of traditional interpolation methods.