首页> 外国专利> Image signal filtering method and apparatus used for extracting scale-invariant feature from image, and image feature extracting method, apparatus and recording medium using thereof

Image signal filtering method and apparatus used for extracting scale-invariant feature from image, and image feature extracting method, apparatus and recording medium using thereof

机译:用于从图像中提取尺度不变特征的图像信号滤波方法和装置,以及使用其的图像特征提取方法,装置和记录介质

摘要

The invention and the image signal filtering method and apparatus used in the constant size of the image feature extraction, using the same image feature extraction method, apparatus, and relates to a recording medium, the image signal filtering method to filter the image signal process according to the invention is used to extract the image feature from the filtered image signal (filtered image signal) by using a size invariant feature extraction algorithm as, SCS (Sparse Code Shrinkage) reduction function (shrinkage function) stabilized reduction function generation step of generating the stabilized reduction function to determine a parameter value of the reduction function from the input image signal to be used in the; Neuroscience image by using the re-expression model generates a sparse matrix and transform coding image signal converting step of generating a conversion value of the input image signal by the sparse coding transformation matrix; Stabilization sparse code generation step of substituting the value of the conversion function to produce a reduction in the stabilization stabilization sparse code of the input image signal; And the sparse and to reverse the stabilized sparse code by the adjoint transformation matrix comprises a filtering of the coded image generation step of generating the filtered image signal, an image, even if there is noise in the input image, and using the RANSAC characterized to reduce the error and to reliably extract.
机译:本发明与图像信号滤波方法和装置一起用于恒定大小的图像特征提取中,使用相同的图像特征提取方法,装置,并且涉及一种记录介质,该图像信号滤波方法用于对图像信号进行滤波处理本发明用于通过使用尺寸不变特征提取算法从滤波后的图像信号(滤波后的图像信号)中提取图像特征,如SCS(稀疏代码收缩)归约函数(收缩函数)的稳定归约函数的生成步​​骤。稳定的缩小函数,用于从要在其中使用的输入图像信号确定缩小函数的参数值;神经科学图像通过使用重新表达模型生成稀疏矩阵并进行变换编码图像信号转换步骤,通过稀疏编码变换矩阵生成输入图像信号的转换值;稳定稀疏码生成步骤,用转换函数的值代入,以减小输入图像信号的稳定稳定稀疏码;并且通过伴随变换矩阵进行的稀疏和反转稳定的稀疏代码包括对编码图像生成步骤的滤波,该步骤生成滤波后的图像信号,即使输入图像中存在噪声的图像,并使用RANSAC减少错误并可靠地提取。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号