# Convolution in Image Processing

Does anyone know how to write arg min code in matlab with explanation?

I am referring to a paper for my research. I am not sure whether I can paste the paper link here. At first, I try to do feature extraction of finger using Gabor Filter. According to the paper, after feature extraction level, I need to do the following using formula to get my desired image:

OriCode(x, y) = arg j min {I(x, y) ∗ G(x, y, θj)}

∗ stands for convolution, I(x,y) stands for input image, G(x,y,θj) stands for gabor filtered image

There is actually a small j between arg and min. OriCode ranges from 0 to J-1 whereby j ranges from 0 to J. Here,J is set to 6.

I try to use conv2 to convolute my original input image and gabor filtered image. End up I am getting a blank white image. And what is the arg min used for? Please help me. Thanks.

• Please post a reference to the paper. – porten Dec 9 '13 at 17:16

I = randn(256, 256); % Just some random image