Convolutional networks: how to deal with pairs of similar filters?

I have to train a convolutional network. When I visualize its conv filter matrices from a layer, many filters are quite similar, almost equal.

1) Don't you know a common way to detect couples of similar kernels? Maybe, it is somehow connected with their correlation?

2) Don't you know any good strategies to deal with such duplicate kernels? My only offer is to re-initialize such duplicate with noise and continue the training, but I don't think it is the best idea.