Why do emgu images have arrays with extra elements for height?

Just noticed something strange and wondered if there was a reason for it.

I have a grayscale image in EmguCV (.net wrapper for openCV). If this image has C channels (1 for grayscale, 3 for RGB), a width of X pixels and a height of Y pixels, the data array is co开发者_Go百科nsistently of the form:

[Y,X+2,C]

Anyone know why?

EmguCV wrapping Opencv relies on its optimization. Opencv's IplImage requires each row of the data matrix to be aligned by 4 bytes. This is done mainly because of the SSE optimization in OpenCV and to optimize data fetching. SSE optimization is one of the technique that make opencv run fast.