玖[9]，带校正的可变形匹配，函数CreatePlanarCalibDeformableModel/FindPlanarCalibDeformableModel

函数CreatePlanarCalibDeformableModel

函数功能

创建一个（带校正透视的）可变形模板

C++形式

LIntExport void CreatePlanarCalibDeformableModel(
const HObject& Template, 
const HTuple& CamParam, 
const HTuple& ReferencePose, 
const HTuple& NumLevels, 
const HTuple& AngleStart, 
const HTuple& AngleExtent, 
const HTuple& AngleStep, 
const HTuple& ScaleRMin, 
const HTuple& ScaleRMax, 
const HTuple& ScaleRStep, 
const HTuple& ScaleCMin, 
const HTuple& ScaleCMax, 
const HTuple& ScaleCStep, 
const HTuple& Optimization, 
const HTuple& Metric, 
const HTuple& Contrast, 
const HTuple& MinContrast, 
const HTuple& GenParamName, 
const HTuple& GenParamValue, 
HTuple* ModelID);


//参数1:模板图像
//参数2:相机标定参数
//参数3:相机标定参考姿态参数
//参数4:金字塔层数，默认"auto"
//参数5:匹配时起始角度
//参数6:匹配时角度范围
//参数7:匹配时旋转角度的步长，默认"auto",参考值>=0
//参数8:Row方向最小缩放比列
//参数9:Row方向最大缩放比列
//参数10:Row方向缩放比列步长
//参数11:Column方向最小缩放比列
//参数12:Column方向最大缩放比列
//参数13:Column方向缩放比列步长
//参数14:模板优化/模板创建方法,默认"auto"
//参数15:模板匹配的方法,默认"use_polarity"
//参数16:对比度增强系数
//参数17:最小对比度阈值,默认"auto",参考值 < 参数8的值
//参数18:输入一般参数名称,默认"",参考值"","part_size"
//参数19:输入一般参数名称,默认"",参考值"","small","medium","big"
//参数20:输出的模板ID

//参数14的参考值
//"auto"
//"none"
//"point_reduction_low",减少点数
//"point_reduction_medium",减少点数
//"point_reduction_high",减少点数
//"pregenneration",预先存储到内存，会快
//"no_pregenneration",预先存储到内存，会慢
//参数15参考值
//"use_polarity",使用极性
//"ignore_global_polarity",忽略全局极性
//"ignore_local_polarity",忽略局部极性
//"ignore_color_polarity",忽略颜色极性

C#形式

public static void CreatePlanarCalibDeformableModel(
HObject template, 
HTuple camParam, 
HTuple referencePose, 
HTuple numLevels, 
HTuple angleStart, 
HTuple angleExtent, 
HTuple angleStep, 
HTuple scaleRMin, 
HTuple scaleRMax, 
HTuple scaleRStep, 
HTuple scaleCMin, 
HTuple scaleCMax, 
HTuple scaleCStep, 
HTuple optimization, 
HTuple metric, 
HTuple contrast, 
HTuple minContrast, 
HTuple genParamName, 
HTuple genParamValue, 
out HTuple modelID);

函数FindPlanarCalibDeformableModel

函数功能

找出图像中一个带校正的可变形模型的最佳匹配

C++形式

LIntExport void FindPlanarCalibDeformableModel(
const HObject& Image, 
const HTuple& ModelID, 
const HTuple& AngleStart, 
const HTuple& AngleExtent, 
const HTuple& ScaleRMin, 
const HTuple& ScaleRMax, 
const HTuple& ScaleCMin, 
const HTuple& ScaleCMax, 
const HTuple& MinScore, 
const HTuple& NumMatches, 
const HTuple& MaxOverlap, 
const HTuple& NumLevels, 
const HTuple& Greediness, 
const HTuple& GenParamName, 
const HTuple& GenParamValue, 
HTuple* Pose, 
HTuple* CovPose, 
HTuple* Score);

//参数1:输入图像
//参数2:模板ID
//参数3:匹配时的起始角度
//参数4:匹配时的角度范围
//参数5:Row方向最小缩放比例
//参数6:Row方向最大缩放比例
//参数7:Column方向最小缩放比例
//参数8:Column方向最大缩放比例
//参数9:匹配时最小得分,默认=0.5
//参数10:匹配时的最大数量,默认=1
//参数11:多个匹配结果的重叠系数,默认=0.5,值=0，不能重叠，值=1，所有匹配结果，参考值>=0 && <=1
//参数12:匹配时，搜索金字塔层数，默认值0，参考值0-10
//参数13:贪婪程度，默认0.9，参考值>=0 && <=1
//参数14:输入参数名称
//参数15:输入参数值
//参数16:输出模板可变形姿态
//参数16:输出Pose参数的6个标准偏差或36个协方差
//参数17:输出得分

参数14参考值
//""
//"subpixel",亚像素
//"max_angle_distortion",最大角度失真
//"angle_step",角度步长
//"max_aniso_scale_distortion",最大比例失真
//"scale_r_step",Row方向比例步长
//"scale_c_step",Column方向比例步长
//"cov_pose_mode",姿态模式

参数15参考值
//""
//"none"
//"least_squares"
//"least_squares_high"
//"least_squares_very_high"

C#形式

public static void FindPlanarCalibDeformableModel(
HObject image, 
HTuple modelID, 
HTuple angleStart, 
HTuple angleExtent, 
HTuple scaleRMin, 
HTuple scaleRMax, 
HTuple scaleCMin, 
HTuple scaleCMax, 
HTuple minScore, 
HTuple numMatches, 
HTuple maxOverlap, 
HTuple numLevels, 
HTuple greediness, 
HTuple genParamName, 
HTuple genParamValue, 
out HTuple pose, 
out HTuple covPose, 
out HTuple score)

附注

线性可变形模板用于在行列方向上可以进行适当的缩放。

函数ReadCamPar

函数ReadPose