If you know GIMP or Photoshop, or some similar image manipulation tool, and know your way around color spaces the way a fish knows it way in water, a hopefully easy way is to decompose the image in LAB images. Use the curve on the B channel, strengthening the negative values. Compose into a color image.
An alternative: decompose to HSV and also to RGB. By whatever means, make a mask where B is greater than R or G. Use that on the S channel image, increasing S only in those areas. Then put the HSV back together into a color image.
If you don't know Photoshop or GIMP, or know nothing of color spaces, learning this can make for an interesting afternoon.
For an in-camera way, one trick I had fun with some years ago, is get some filters of various colors, pale amber, pink, pale green, violet, etc. Not just those filters for dealing with indoor/outdoor or florescent/incandescent situations. Get interesting colors, a variety. Put a color filter on the lens, aim your camera at something white, and tell your camera to do a white balance. Take photos. White is still white, black is black, but colorful objects will look different than normal for that camera. For more vibrant blue, it's more trial and error than reasoning, since characteristics of the camera and its internal processing of color make it hard to know just how a photo will turn out. Experiment.