In this paper, we propose a new shape-coding algorithm called wavelet-based shape coding (WBSC). Performing wavelet transform on the orientation of original planar curve gives the corners called corner-1 points and en...In this paper, we propose a new shape-coding algorithm called wavelet-based shape coding (WBSC). Performing wavelet transform on the orientation of original planar curve gives the corners called corner-1 points and end of arcs that belong to the original curve. Each arc is represented by a broken line and the corners called corner-2 points of the broken line are extracted. A polygonal approximation of a contour is an ordered list of corner-1 points, ends of arcs and corner-2 points which are extracted by using the above algorithm. All of the points are called polygonal vertices which will be compressed by our adaptive arithmetic encoding. Experimental results show that our method reduces code bits by about 26% compared with the context-based arithmetic encoding (CAE) of MPEG-4, and the subjective quality of the reconstructed shape is better than that of CAE at the same Dn.展开更多
文摘In this paper, we propose a new shape-coding algorithm called wavelet-based shape coding (WBSC). Performing wavelet transform on the orientation of original planar curve gives the corners called corner-1 points and end of arcs that belong to the original curve. Each arc is represented by a broken line and the corners called corner-2 points of the broken line are extracted. A polygonal approximation of a contour is an ordered list of corner-1 points, ends of arcs and corner-2 points which are extracted by using the above algorithm. All of the points are called polygonal vertices which will be compressed by our adaptive arithmetic encoding. Experimental results show that our method reduces code bits by about 26% compared with the context-based arithmetic encoding (CAE) of MPEG-4, and the subjective quality of the reconstructed shape is better than that of CAE at the same Dn.
基金the National Natural Science Foundation of China under Grant Nos.60607010,60672125,60672126,60736043,60776795(国家自然科学基金)the Program for Cheung Kong Scholars and Innovative Research Team in University of China under Grant No.IRT0645(教育部长江学者和创新团队支持计划)the Natural Science Foundation of Hu'nan Province of China under Grant No.08JJ3123(湖南省自然科学基金)
文摘提出了一种针对三维医学图像已知诊断对象区域的形状自适应小波编码方法.该算法仅对对象区域内的像素应用形状自适应小波变换去相关,变换后,对象在变换域中的系数个数与图像域的像素保持相同.为了实现快速无损变换,提出一种基于提升的形状自适应小波变换方法.通过分析形状自适应变换后无效系数的位置,又提出一种改进的OB-3DSPECK(object-based set partitioned embedded block coder)算法,取消了对象区域外无效块或系数的符号输出,即只输出两种符号码流到自适应算术编码器.对于三维医学图像的对象区域,该算法能够提供有损到无损的渐进编解码.实验结果表明,该算法平均SNR比OB-3DSPECK提高0.5dB.此外,由于减少了一种符号输出,使得算术编码过程可选.
