cv::Mat 取元素引起的报错

一、定义一个矩阵

方法1：定义的时候，矩阵元素无初始值，需要赋值

cv::Mat intrinsic(3, 3, CV_64F );

cv::Mat distortion(1, 5, CV_64F);

1.矩阵赋值

intrinsic.at<double>(0, 0) = 2400;

intrinsic.at<double>(0, 1) = 0;

intrinsic.at<double>(0, 2) = 960;

intrinsic.at<double>(1, 0) = 0;

intrinsic.at<double>(1, 1) = 2400;

intrinsic.at<double>(1, 2) = 540;

intrinsic.at<double>(2, 0) = 0;

intrinsic.at<double>(2, 1) = 0;

intrinsic.at<double>(2, 2) = 1;

distortion.at<double>(0, 0) = 0 ;

distortion.at<double>(0, 1) = 0;

distortion.at<double>(0, 2) = 0;

distortion.at<double>(0, 3) = 0;

distortion.at<double>(0, 4) = 0;

方法2：定义的时候，矩阵元素初始为0

cv::Mat intrinsic = cv::Mat::zeros(3, 3, CV_64FC1);

cv::Mat distortion = cv::Mat::zeros(1, 5, CV_64FC1);

intrinsic.at<double>(0, 0) = 2400;

intrinsic.at<double>(0, 2) = 960;

intrinsic.at<double>(1, 1) = 2400;

intrinsic.at<double>(1, 2) = 540;

intrinsic.at<double>(2, 2) = 1;

方法3：cv::Mat_ 定义矩阵，直接初始化

cv::Mat distortion = (cv::Mat_<double>(1, 5) << 0, 0, 0, 0, 0);

二、定义的Mat类型，与使用时的类型不一致

使用Mat .at时数据类型的选择错误：如double类型的数据，你写成了float就会报错。

intrinsic.at<float>(0, 0) = 2000;

报错1：OpenCV(4.4.0) Error: Assertion failed (((0x28442211 >> ((traits::Depth<_Tp>::value) & ((1 << 3) - 1))*4) & 15) == elemSize1()) in cv::Mat::at, file D:\Projects\C++\opencv_test\opencv_test\opencv_test\include\opencv2\core\mat.inl.hpp, line 1157

三、Mat.at 取值越界

intrinsic.at<double>(3, 0) = 2000;

报错2：OpenCV(4.4.0) Error: Assertion failed (data) in cv::Mat::at, file D:\Projects\C++\opencv_test\opencv_test\opencv_test\include\opencv2\core\mat.inl.hpp, line 1154