[SOLVED] No instance of overloaded function "pcl::eigen33" matching the argument list

mistake	C2664	'void pcl::eigen33<Eigen::Matrix3f,Eigen::Vector3d>(const Matrix &,float &,Vector &)': cannot convert argument 2 from 'Eigen::Matrix3d' to 'float &'			3DVisionMFCshf		F:\workcplus\3DVisionMFCshf\3DVisionMFCshf\3DVisionMFCshf.cpp	3981
severity	the code	illustrate	project	document	OK	Disable status
mistake	C2338	YOU_MIXED_DIFFERENT_NUMERIC_TYPES__YOU_NEED_TO_USE_THE_CAST_METHOD_OF_MATRIXBASE_TO_CAST_NUMERIC_TYPES_EXPLICITLY	3DVisionMFCshf	C:\pcl12\PCL 1.12.1\3rdParty\Eigen\eigen3\include\eigen3\Eigen\src\Core\AssignEvaluator.h	834

 

 I don’t know why the code in the above picture reported an error. After the computer system was reinstalled and all configurations were completed, only this line of the project reported an error. In the original system, this line of code was fine, and the specific reason was not found.

 

 The solution is to modify the above code and change it to the following code to solve the problem

Eigen::Matrix3f eigenVectors;
	Eigen::Vector3f eigenValues;
	pcl::eigen33<Eigen::Matrix3f, Eigen::Vector3f>(covariance, eigenVectors, eigenValues);

This algorithm is mainly used when calculating the covariance matrix.