shahidhk / cv2cg Goto Github PK

in CV_CG.h:
As you mentioned, we don't need (x, y, w, h) to do cv2cg, however, you 
preserved. That needs correct n, yet, not every .par file provides n and the 
default value is 1.0, which will result in enormously unbalanced r and l.

IMHO, delete all the ImgW and ImgH. 

I've use the DLTCalibrator to do some test, using numbers(>6) of points 
correspondencies, world point (X,Y,Z) in mms, image point (u, v) in pixels, the 
output camera matrix seems wrong, fx/fy/u0/v0 are less than 1, i find the 
algorithm doesn't need the image size (width, height) as an parameter, is that 
the reason or i use the wrong units? can you explain how to use DLTCalibrator 
to get the intrinsic and extrinsic matrix correctly given the 3d-2d points? 
thank you very much.

Hello,
in the main CMakeLists.txt there is a ADD_SUBDIRECTORY(data) command.

However, there is no CMakeLists.txt in the data directory, so the compilation 
fails.

Thanks a lot,
Christian

目前的条件编译帮助字符串的语言采用的是USE_IN_CHINA宏，这��
�可以考虑把字符串放入资源中，然后使用字符串的时候调用�
��是资源，这个可以根据Windows的语言系统来自动选择的。