基于多曝光场景的新型电力图像配准技术

New Power Image Registration Technology Based on Multi-exposure Scene

基于现有的图像配准技术,对电力图像拍摄时间差异所造成的曝光度不同而产生的图像配准误差大和配准效率提升难的问题,提出了一种基于电力图像去除亮度差异影响的边缘特征配准技术。通过提取图像的边缘信息,以产生二值化的效果图像代替原始图像进行旋转矩阵的计算,完成原始图像空间一致性调整,实现多曝光场景的新型电力图像配准技术,并通过一系列电力图像试验数据进行了验证。试验结果表明,所提的新型电力图像配准技术在配准效率方面较改进前的基于互相关信息配准方法提升了50%,配准精度达到了亚像素级别,配准方法鲁棒性更高。

Based on the problems of large image registration error due to different exposure caused by the difference of power image shooting time and difficult to improve the registration efficiency,an edge feature registration technology based on removing the influence of brightness difference of power image was proposed.By extracting the edge information of the image,it can generate a binary effect image instead of the original image to calculate the rotation matrix,complete the original image spatial consistency adjustment,and realize the new power image registration technology for multi-exposure scenes.A series of electric power image experimental data were verified.The test results show that the proposed new power image registration technology has improved the registration efficiency by 50%compared with the registration method based on cross-correlation information before the improvement,the registration accuracy has reached the sub-pixel level,and the registration method is of higher robustness.