期刊文献+

彩色与纹理不变性的阴影消除新算法 被引量:3

New algorithm to eliminate shadows for the invariance of the color and texture
下载PDF
导出
摘要 针对阴影C1C2C3彩色不变性检测法时间开销大的缺陷,提出了改进的D1D2D3模型阴影检测法,减少了算法的时间复杂度;针对阴影LBP纹理不变性检测法在图像纹理不明显时所存在的误检现象,提出了改进的LPTD算子,提高了算法的准确率。进一步将上述两种改进方法进行有效融合,提出了一种阴影消除新算法,采用SUSAN算子去除阴影强边缘,有效解决了强光给阴影检测带来的误差,再通过形态学处理,消除了图像内部孔洞和噪声。实验结果表明,新算法的阴影消除效果更为理想。 According to defects existing in time complexity in the shadow model of the C1C2C3 which uses invariant color features, a method of DID2D3 shadow model is proposed, which can reduce the time complexity of the algorithm. According to defects existing in mistake detection in shadow LBP using invariant texture method when the image texture is not evident, the LPTD operator is put forward. And by fusing the two improved methods effectively, a new algorithm to eliminate shadows is proposed, which uses the SUSAN operator to remove the strong edge of shadow arid effectively solves the error caused by the strong light in the shadow eliminating. Through the morphological processing, the internal holes and noise of image are eliminated. Experimenl results show that the new algorithm is more ideal in the effect of shadow eliminating.
出处 《微型机与应用》 2013年第5期38-41,45,共5页 Microcomputer & Its Applications
基金 国家自然科学基金(51277032)
关键词 彩色不变性 纹理不变性 阴影消除 SUSAN the invariance of the color the invariance of the texture shadow eliminate SUSAN
  • 相关文献

参考文献6

二级参考文献55

共引文献32

同被引文献26

  • 1徐建瑜,崔绍荣,苗香雯,刘鹰.计算机视觉技术在水产养殖中的应用与展望[J].农业工程学报,2005,21(8):174-178. 被引量:39
  • 2Sanders J,Kandrot E.GPU高性能编程CUDA实战[M].聂雪军,译.北京:机械工业出版社,2011.
  • 3KIRKDB,HWUWW.大规模并行处理器编程实战[M].陈曙晖,熊淑华,译.北京:清华大学出版社.2010.
  • 4AHONEN T, HADID A, PIETIKAINEN M. Face description with local binary pattern: application to face recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 28(12):2037-2041.
  • 5OJALA T, PIETIKAINEN M, MAENPAA T. Muhiresolution gray-scale and rotation-invariant texture classification with local binary patterns[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002, 24(7):971-987.
  • 6LEE S H, CHOI J Y, ROY M, et al. Local color vector binary patterns from multi-channel face images for face recognition[J]. IEEE Transactions on Image Processing, 2012, 21 (4) :2347-2353.
  • 7OJALA T, MAENPAA T, PIETIKAINEN M, et al. Outex-a new framework for empirical evaluation of texture analysis algorithms[C]. Proceedings of the 16th Conference on Pattern Recognition, 2002:701-706.
  • 8郝灿,朱信忠,赵建民,徐慧英.基于改进型LBP特征的运动阴影去除算法[J].计算机系统应用,2010,19(5):80-83. 被引量:7
  • 9余孟泽,刘正熙,骆键,杨丹.融合纹理特征和阴影属性的阴影检测方法[J].计算机工程与设计,2011,32(10):3431-3434. 被引量:12
  • 10曹健,陈红倩,张凯,牛长锋.结合区域颜色和纹理的运动阴影检测方法[J].机器人,2011,33(5):628-633. 被引量:17

引证文献3

二级引证文献11

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部