2D Convolution for Image Processing

preview_player
Показать описание
2D convolution animation video image processing
Рекомендации по теме
Комментарии
Автор

thank you, helped to understand the concept

RicardoLizarraga-ufzn
Автор

if the kernel wasn't symmetric, wouldn't we fold (flip in both dimensions) the kernel before multiplying at each and every cell?

mr.crazyunknown
Автор

Why does the kernel is multiplied with its middle point on the top left corner, and the size of output matrix should be 3x2, instead of being same, could you please clear

rishabhtibrewal
Автор

what does it mean when there is a underscored cell in signal and filter

kerimhasanyildirim
Автор

Very nice and explanatory example. Yet, this seems to me not the ordinary convolution. Why does the size of the matrix remain the same? If you answer, I would be glad

yarenkaya
Автор

For example if h=[1 1 1;0 0 0;-1 -1 -1], after reflextion on x and y coordinates, it becomes [-1 -1 -1;0 0 0;1 1 1]. If h=[1 0 -1;1 0 -1;1 0 -1], it becoms [-1 0 1;-1 0 1;-1 0 1]

onurosman