基于参数自适应滤波器的抠图优化算法

摘要

In order to solve when using guided image filter algorithms for image matting, the parameters need to manually adjust according to the different image. A method was proposed to adaptive parameters. This approach draws on the idea of image evaluation, after divided image with different radius, using the average gradient value of each block to determine the radius, so as to achieve the purpose of automatically generated reasonable parameters. While matting experiments, using this adaptive filter optimization masking value in the digital image matting. Experimental results show that the adaptive filtering algorithm can capture the image detail very well, reached adaptation purpose.%针对应用目标图像滤波算法进行图像抠图时,参数需要根据不同图像进行手动调整的问题,提出了一种使其参数自适应的方法.该方法借鉴了图像评价的思路,将图像以不同半径分块后,利用图像各个分块值的平均梯度变化,确定半径值,从而达到了自动产生合理参数的目的.同时在抠图实验中,运用该自适应滤波算法对数字抠图中的掩膜值进行计算优化.实验结果表明自适应化后的滤波算法能够很好地捕捉图像中的细节,达到自适应的目的.