基于内容的压缩域图像处理技术研究

研究了当前压缩域图像处理技术,针对目前小波技术在图像编码标准中的广泛运用,提出了一种结合数学形态学技术直接在小波域内进行图像轮廓提取的新方法.其思想是依据图像的小波编码码流元素的幅度与图像轮廓在频率-空间上具有的对应关系,利用数学形态学中以具有一定形态的结构元素去量度和提取图像中的对应形状,以快速提取出反映图像基本内容的轮廓特征.对小波域中的高频系数进行处理,因此首先要对小波图像做阈值化,然后采用腐蚀膨胀算法做去噪和轮廓提取.实验结果表明,与微分算子提取图像轮廓相比,计算简单.易于硬件实现,具有良好的性能.

An Image Processing Technique in Compressed Domain Based on Content

The paper mainly studies the image processing technique in compressed domain.In view of the wide application of wavelet in coding and encoding standard,the paper proposes a figure pick-up technique,which utilizes mathematical morphology.The idea is that the wavelet coefficients correspond to image's figure,and the mathematical morphology utilizes framework element to measure and pick-up figures.This technique can retrieve the target image based on content fleetly.The technique handles high frequency coeffic...