D-SIFT一种面向DCT域的扩展SIFT特征.pdfVIP

第27 卷 第 10 期 计算机辅助设计与图形学学报 Vol. 27 No.10 2015 年 10 月 Journal of Computer-Aided Design Computer Graphics Oct. 2015 D-SIFT ：一种面向DCT 域的扩展SIFT 特征 庄连生, 费 驰, 钟航世, 俞能海 1) ( 中国科学技术大学信息学院 合肥 230027) 2) ( 中国科学院电磁空间信息重点实验室 合肥 230027) (lszhuang@ustc.edu.cn) : 提出一种面向DCT 域的扩展 SIFT 特征(即D-SIFT 特征), 以提高JPEG 图像的匹配速度. D-SIFT 特征首先 在 DCT 域对图像进行不同尺度的高斯滤波, 构建出图像的尺度空间; 然后, 把滤波后的图像划分为大小为 2×2 的图 像块, 并通过检测出极值图像块来估计关键特征点位置; 最后, 利用当前块周围 16×16 个子块的 DCT 直流系数来统 计主方向, 提取每个极值块的描述子. 实验结果表明, D-SIFT 特征可以有效地提高JPEG 图像的匹配效率. ：压缩域图像处理; JPEG 图像匹配; SIFT 特征提取 ：TP391.41 D-SIFT: an Extended SIFT Feature in the DCT Domain Zhuang Liansheng, Fei Chi, Zhong Hangshi, and Yu Nenghai 1) (School of Information Science and Technology, University of Science and Technology of China, Hefei 230027) 2) (CAS Key Laboratory of Electromagnetic Space Information, Hefei 230027) Abstract: To accelerate the speed of JPEG image matching, we propose a novel local feature called D-SIFT feature, which is an extension of SIFT in the DCT domain. Firstly, D-SIFT does Gaussian filtering to the image in the DCT domain, obtaining the image scale space. Then, D-SIFT divides all the filtered images into 2×2 image blocks, and estimates all the key points by detecting the key blocks. Finally, D-SIFT obtains the main orientation and extracts the descriptor for each key block from its surrounding 16×16 blocks by using their DC coefficients. Experimental results show that, D-SIFT feature can efficiently improve the speed of JPEG image matching. Key words: image processing in compressed domain; JPEG image matching; SIFT feature extraction [5] . (), 1 相关工作 .