摘要
针对分形图像编码时间长的问题,提出选取边缘提取图像的6个特征参数作为图像块特征.采用模糊模式识别技术对图像块进行分类,后采用基于局部灰度均值聚类技术进一步减小最佳匹配搜索范围的两步分类法.实验表明,该方法可较大提高编码速度,解码质量无明显下降.
With the problem of expensive time spending in the fractal image encoding, an approach is proposed that 6 parameters of edge image blocks are selected as the features of image blocks. The original image blocks are classified by a fuzzy pattern recognition technology. Then, the cluster analysis technology based on part grey means is used to shrink the search ranges for optimal matching block. The simulation results show that the approach can speed up encoding time greatly with no apparent deterioration on the quality of the reconstructed image.
出处
《计算机辅助工程》
2006年第3期79-82,共4页
Computer Aided Engineering
关键词
分形
图像压缩
模糊分类
聚类分析
fractal
image compression
fuzzy classification
cluster analysis
