Optimal target location in multichannel images

We propose algorithms for the location of a known target in a multichannel image, optimal for the multichannel model of the image, and optimal for the noise model of the image. We consider two noise models: Gaussian additive, and Gaussian non-overlapping. We show on simulations the improvement due to the multichannel model of the image, compared with the monochannel case. For simulations, the multichannel images are color images.