I remember writing code for convolutions in CS61C last semester. We used it to create a blur filter over a gif of Lightning McQueen. I thought it was really cool to see the conceptual connection in this class.
jayc809
The commutative and associative nature of the convolution operation on the spatial domain suggests that we can combine two convolution filters by taking their sum. In other words, instead of convolving the image with one filter and then convolving that result with the other filter, we can instead sum the two filters first and then convolve that filter with the image. This could save a lot of computation when there are multiple filters involved and potentially be a lot more efficient for larger images.
I remember writing code for convolutions in CS61C last semester. We used it to create a blur filter over a gif of Lightning McQueen. I thought it was really cool to see the conceptual connection in this class.
The commutative and associative nature of the convolution operation on the spatial domain suggests that we can combine two convolution filters by taking their sum. In other words, instead of convolving the image with one filter and then convolving that result with the other filter, we can instead sum the two filters first and then convolve that filter with the image. This could save a lot of computation when there are multiple filters involved and potentially be a lot more efficient for larger images.