基于色彩校正与分层块匹配的图像配准算法
|
摘要
针对当前图像匹配算法难以兼顾匹配精度和效率,易受模糊、噪声和色彩不一致影响等问题,提出色彩校正耦合分层块匹配优化的彩色图像配准算法。计算RGB图像每个色彩通道强度的平均值和标准差,建立色彩映射方程,对色彩不一致进行补偿,完成色彩校正;基于广义高斯分布金字塔RGB强度的差分和,确定高斯金字塔层数,进行特征检测,计算特征位置;进行初始块匹配,利用变化的单位比例因子对计算成本优化,利用抛物线拟合技术处理金字塔的每个通道,寻找其亚像素精度的最优匹配。仿真结果表明,该算法能够提高计算效率和精度,在色彩不一致、模糊和噪声图像上取得了良好配准结果,具有较强抗干扰能力与较高的鲁棒性。
The current image matching algorithm is difficult to give attention to matching accuracy and efficiency,and it is easy to be influenced by the image blur,noise and color.To solve the problem,the color image registration algorithm based on color registration coupled layer block matching was proposed.The average value and standard deviation of each color channel intensity were calculated.The color mapping equation was built by means of draw value and standard deviation,color mapping equation was established,so the color correction was completed.The color correction was performed.Based on the generalized Gauss distribution of pyramid RGB,the feature positions were determined.The initial block matching was performed.The change unit scale factor was used to process calculation cost,parabolic fitting technology was used to process each channel of pyramid,and the optimal matching of the sub pixel accuracy was calculated using the variation of RGB.Simulation results show that the proposed algorithm can improve the efficiency and accuracy of image registration,and it has strong anti-interference ability and robust performance.
The current image matching algorithm is difficult to give attention to matching accuracy and efficiency,and it is easy to be influenced by the image blur,noise and color.To solve the problem,the color image registration algorithm based on color registration coupled layer block matching was proposed.The average value and standard deviation of each color channel intensity were calculated.The color mapping equation was built by means of draw value and standard deviation,color mapping equation was established,so the color correction was completed.The color correction was performed.Based on the generalized Gauss distribution of pyramid RGB,the feature positions were determined.The initial block matching was performed.The change unit scale factor was used to process calculation cost,parabolic fitting technology was used to process each channel of pyramid,and the optimal matching of the sub pixel accuracy was calculated using the variation of RGB.Simulation results show that the proposed algorithm can improve the efficiency and accuracy of image registration,and it has strong anti-interference ability and robust performance.
引文
[1]DING Nannan,LIU Yanying,ZHANG Ye.Fast image registration based on SURF-DAISY algorithm and ran-demised kd tree[J].Journal of Optoelectronics·Laser,2012,23(7):1395-1402(in Chinese).[丁南南,刘艳滢,张叶.基于SURF-DAISY算法和随机kd树的快速图像配准[J].光电子·激光,2012,23(7):1395-1402.]
[2]HAN Yu,WANG Weiwei,FENG Xiangchu.Iteratively reweighed method based nonrigid image registration[J].Acta Automatica Sinica,2011,37(9):1059-1066(in Chinese).[韩雨,王卫卫,冯象初.基于迭代重加权的非刚性图像配准[J].自动化学报,2011,37(9):1059-1066.]
[3]Andreas D Merrem,Frank G Zllner,Marcel Reich.A variational approach to image registration in dynamic contrast-enhanced MRI of the human kidney[J].Magnetic Resonance Imaging,2013,31(5):771-778.
[4]WANG Li,ZENG Peipei,HAO Jianxin.Infrared image registration of circuit board based on ROI gray compression[J].Laser&Infrared,2014,44(3):313-318(in Chinese).[王力,曾佩佩,郝建新.基于ROI灰度压缩的电路板红外图像配准[J].激光与红外,2014,44(3):313-318.]
[5]XU Liyan.Research on remote sensing image registration algorithm based on feature points and application[D].Nanjing:Nanjing University of Science and Technology,2012:20-41(in Chinese).[徐丽燕.基于特征点的遥感图像配准方法及应用研究[D].南京:南京理工大学,2012:20-41.]
[6]Lu Xiaoqi,Zhao Yongjie,Zhang Baohua.A non-rigid cardiac image registration method based on an optical flow model[J].Optik,2013,124(20):4266-4273.
[7]XU Chen.Research and application of image recognition method based on block matching[D].Beijing:North China University of Technology,2015:22-28(in Chinese).[徐晨.基于块匹配的图像识别方法研究及应用[D].北京:北方工业大学,2015:22-28.]
[8]LI Xiaoyun,DU Wei.Analysis of the influence of image motion on the image quality of TDICCD camera[J].Spacecraft Engineering,2011,20(3):51-55(in Chinese).[李晓云,杜伟.星载TDICCD相机像移对成像质量的影响分析[J].航天器工程,2011,20(3):51-55.]
[9]Wang Xiangyang,Zhao Li,Niu Panpan.Image denoising using gaussian scale mixtures with gaussian-hermite PDF in steerable pyramid domain[J].Journal of Mathematical Imaging and Vision,2011,39(3):245-258.
[10]CHEN Li,LONG Guangli.Application of Retinex algorithm in face recognition based on sparse difference and Mean-Shift filtering[J].Application Research of Computers,2015,32(3):934-937(in Chinese).[陈莉,龙光利.基于稀疏差分和Mean-Shift滤波的Retinex算法在人脸识别中的应用[J].计算机应用研究,2015,32(3):934-937.]
[11]ZHONG Ying,YANG Xuezhi,TANG Yiming.Non-local means denoising derived from structure-adapted block matching[J].Journal of Electronics&Information Technology,2013,35(12):2908-2915(in Chinese).[钟莹,杨学志,唐益明.采用结构自适应块匹配的非局部均值去噪算法[J].电子与信息学报,2013,35(12):2908-2915.]
[12]MENG Weican,ZHU Shulong,CAO Wen.Point feature extraction of remote sensing image using SI-Harris[J].Journal of Geomatics Science and Technology,2014,31(4):383-387(in Chinese).[孟伟灿,朱述龙,曹闻.运用SI-Harris算子提取遥感图像点特征[J].测绘科学技术学报,2014,31(4):383-387.]
[13]Changsoo Jen,Park Hyung-Min.Optimized hierarchical block matching for fast and accurate image registration[J].Image Communication,2013,28(7):779-791.
[2]HAN Yu,WANG Weiwei,FENG Xiangchu.Iteratively reweighed method based nonrigid image registration[J].Acta Automatica Sinica,2011,37(9):1059-1066(in Chinese).[韩雨,王卫卫,冯象初.基于迭代重加权的非刚性图像配准[J].自动化学报,2011,37(9):1059-1066.]
[3]Andreas D Merrem,Frank G Zllner,Marcel Reich.A variational approach to image registration in dynamic contrast-enhanced MRI of the human kidney[J].Magnetic Resonance Imaging,2013,31(5):771-778.
[4]WANG Li,ZENG Peipei,HAO Jianxin.Infrared image registration of circuit board based on ROI gray compression[J].Laser&Infrared,2014,44(3):313-318(in Chinese).[王力,曾佩佩,郝建新.基于ROI灰度压缩的电路板红外图像配准[J].激光与红外,2014,44(3):313-318.]
[5]XU Liyan.Research on remote sensing image registration algorithm based on feature points and application[D].Nanjing:Nanjing University of Science and Technology,2012:20-41(in Chinese).[徐丽燕.基于特征点的遥感图像配准方法及应用研究[D].南京:南京理工大学,2012:20-41.]
[6]Lu Xiaoqi,Zhao Yongjie,Zhang Baohua.A non-rigid cardiac image registration method based on an optical flow model[J].Optik,2013,124(20):4266-4273.
[7]XU Chen.Research and application of image recognition method based on block matching[D].Beijing:North China University of Technology,2015:22-28(in Chinese).[徐晨.基于块匹配的图像识别方法研究及应用[D].北京:北方工业大学,2015:22-28.]
[8]LI Xiaoyun,DU Wei.Analysis of the influence of image motion on the image quality of TDICCD camera[J].Spacecraft Engineering,2011,20(3):51-55(in Chinese).[李晓云,杜伟.星载TDICCD相机像移对成像质量的影响分析[J].航天器工程,2011,20(3):51-55.]
[9]Wang Xiangyang,Zhao Li,Niu Panpan.Image denoising using gaussian scale mixtures with gaussian-hermite PDF in steerable pyramid domain[J].Journal of Mathematical Imaging and Vision,2011,39(3):245-258.
[10]CHEN Li,LONG Guangli.Application of Retinex algorithm in face recognition based on sparse difference and Mean-Shift filtering[J].Application Research of Computers,2015,32(3):934-937(in Chinese).[陈莉,龙光利.基于稀疏差分和Mean-Shift滤波的Retinex算法在人脸识别中的应用[J].计算机应用研究,2015,32(3):934-937.]
[11]ZHONG Ying,YANG Xuezhi,TANG Yiming.Non-local means denoising derived from structure-adapted block matching[J].Journal of Electronics&Information Technology,2013,35(12):2908-2915(in Chinese).[钟莹,杨学志,唐益明.采用结构自适应块匹配的非局部均值去噪算法[J].电子与信息学报,2013,35(12):2908-2915.]
[12]MENG Weican,ZHU Shulong,CAO Wen.Point feature extraction of remote sensing image using SI-Harris[J].Journal of Geomatics Science and Technology,2014,31(4):383-387(in Chinese).[孟伟灿,朱述龙,曹闻.运用SI-Harris算子提取遥感图像点特征[J].测绘科学技术学报,2014,31(4):383-387.]
[13]Changsoo Jen,Park Hyung-Min.Optimized hierarchical block matching for fast and accurate image registration[J].Image Communication,2013,28(7):779-791.