OpenCV：移动单摄像机的旋转矩阵和平移向量 [英] OpenCV: Rotation Matrix and Translation Vector for a Moving Single Camers

问题描述

我有一个校准（内部参数）相机。后来，我需要用它来拍摄物体的不同照片。它们将用于重新投影图像的某些点。

I have a calibrated (intrinsic parameters) camera. Later, I need to use it for taking different photos from an object. They will be used to re-project some points of the images.

关键步骤是使用同一对象的2个图像计算旋转矩阵和平移向量。我是图像处理的新手，我不知道该怎么做。任何信息都将不胜感激。

A key step is to calculate Rotation matrix and Translation vector, using 2 images of a same object. I am a newbie in image processing and I have no idea how to do that. Any information would be appreciated.

谢谢！

推荐答案

1. 在图像A中查找要素点


2. 在图像B中查找要素点


3. 查找要素对应


4. 使用 cv： ：findFundamentalMat 找到基本矩阵。


5. 在E = K'FK找到基本矩阵，其中K是相机内在参数矩阵


6. 现在狡猾部分（被问到之前：正如@Majek建议的那样，使用 https://github.com/MasteringOpenCV/code ，它会给你一个良好的开端。

1. find feature points in image A
2. find feature points in image B
3. find feature correspondence
4. use cv::findFundamentalMat to find the fundamental matrix.
5. find the essential matrix by E = K'FK where K is the camera intrinsics parameters matrix
6. now for the tricky part (that was asked before): as suggested by @Majek, use https://github.com/MasteringOpenCV/code, it will give you a good start.

这篇关于OpenCV：移动单摄像机的旋转矩阵和平移向量的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！