针对旋转不变性二进制描述算法(Oriented Fast and Rotated Brief,ORB)的尺度旋转性配准误差大,配准率较低及随机采样一致性(Random Sample Consensus,RANSAC)算法随机性强且不稳定的问题,提出一种ORB与RANSAC结合的快速特征匹配算法。...针对旋转不变性二进制描述算法(Oriented Fast and Rotated Brief,ORB)的尺度旋转性配准误差大,配准率较低及随机采样一致性(Random Sample Consensus,RANSAC)算法随机性强且不稳定的问题,提出一种ORB与RANSAC结合的快速特征匹配算法。首先,对特征点提取方式进行优化选择,消除特征边缘影响。之后构建简化的金字塔式尺度空间模型,改进分层图像的尺度空间结构,减少生成图像层数和数目;然后采用梯度方向改进传统ORB算法中的主方向提取模式,提高特征角点主方向的准确性。最后,通过构建分块随机取样检测的方式改进RANSAC算法,提高RANSAC算法的稳定性和图像配准的准确性。实验结果表明改进后的ORB和RANSAC融合算法在尺度和旋转配准方面性能有很大提高,并且配准的精度较传统ORB算法高,尺度配准精度提高55.41%,旋转配准精度提高26.66%。满足复杂图像快速精确配准拼接的精度和实时性要求。展开更多
Random needle embroidery(RNE) is a graceful art enrolled in the world intangible cultural heritage. In this paper, we study the stitch layout problem and propose a controllable stitch layout strategy for RNE. Using ou...Random needle embroidery(RNE) is a graceful art enrolled in the world intangible cultural heritage. In this paper, we study the stitch layout problem and propose a controllable stitch layout strategy for RNE. Using our method, a user can easily change the layout styles by adjusting several high-level layout parameters. This approach has three main features: firstly, a stitch layout rule containing low-level stitch attributes and high-level layout parameters is designed; secondly, a stitch neighborhood graph is built for each region to model the spatial relationship among stitches; thirdly, different stitch attributes(orientations, lengths, and colors) are controlled using different reaction-diffusion processes based on a stitch neighborhood graph. Moreover, our method supports the user in changing the stitch orientation layout by drawing guide curves interactively. The experimental results show its capability for reflecting various stitch layout styles and flexibility for user interaction.展开更多
文摘针对旋转不变性二进制描述算法(Oriented Fast and Rotated Brief,ORB)的尺度旋转性配准误差大,配准率较低及随机采样一致性(Random Sample Consensus,RANSAC)算法随机性强且不稳定的问题,提出一种ORB与RANSAC结合的快速特征匹配算法。首先,对特征点提取方式进行优化选择,消除特征边缘影响。之后构建简化的金字塔式尺度空间模型,改进分层图像的尺度空间结构,减少生成图像层数和数目;然后采用梯度方向改进传统ORB算法中的主方向提取模式,提高特征角点主方向的准确性。最后,通过构建分块随机取样检测的方式改进RANSAC算法,提高RANSAC算法的稳定性和图像配准的准确性。实验结果表明改进后的ORB和RANSAC融合算法在尺度和旋转配准方面性能有很大提高,并且配准的精度较传统ORB算法高,尺度配准精度提高55.41%,旋转配准精度提高26.66%。满足复杂图像快速精确配准拼接的精度和实时性要求。
基金Project supported by the National Natural Science Foundation of China(Nos.61272219,61100110,and 61321491)the National High-Tech R&D Program(863)of China(No.2007AA01Z334)+3 种基金the Key Projects Innovation Fund of State Key Laboratory(No.ZZKT2013A12)the Program for New Century Excellent Talents in Universities,China(No.NCET04-04605)the Science and Technology Program of Jiangsu Province(Nos.BE2010072,BE2011058,and BY2012190)the Scientific Research Foundation of Graduate School of Nanjing University(No.2012CL21),China
文摘Random needle embroidery(RNE) is a graceful art enrolled in the world intangible cultural heritage. In this paper, we study the stitch layout problem and propose a controllable stitch layout strategy for RNE. Using our method, a user can easily change the layout styles by adjusting several high-level layout parameters. This approach has three main features: firstly, a stitch layout rule containing low-level stitch attributes and high-level layout parameters is designed; secondly, a stitch neighborhood graph is built for each region to model the spatial relationship among stitches; thirdly, different stitch attributes(orientations, lengths, and colors) are controlled using different reaction-diffusion processes based on a stitch neighborhood graph. Moreover, our method supports the user in changing the stitch orientation layout by drawing guide curves interactively. The experimental results show its capability for reflecting various stitch layout styles and flexibility for user interaction.