Day record - 2007 May 23

Finally, I made it. I found out a little GOD DAMN BUG!!
That is the type of IplImage image data format is "char".That means if it is bigger than 128, while changed to "int", it will appear as a negative one. In computer vision, the range of each pixel's value is 0 ~ 255. In order to process correctly, each pixel value must be converted from (char*) to (unsigned char*) by cast for IplImage format( openCV ). Ex. unsigned char* im = (unsigned char*) IplImage->imageData;