The present paper extends previous results on the sensor correlation method for illuminant classification. We describe several algorithm modifications that improve classification accuracy and applicability to a variety of scenes. First, we use the reciprocal scale of color temperature, called “mired”, in order to obtain perceptually uniform illuminant classification, rather than physical illuminant classification by the original color temperature scale. Second, we introduce a new image normalization operation with an adjustable parameter to adjust overall intensity differences between images and find a good fit to the illuminant gamuts. Third, we calculate a correlation value between an image gamut and the reference illuminant gamut, rather than between the image pixels and the illuminant gamuts. This calculation makes it more reliable to select a unique illuminant gamut as an estimate. Fourth, we develop the 3D classification algorithms using all three-color channels, and compare this with the original 2D algorithms from the viewpoint of accuracy and efficiency. Finally, the improved algorithms incorporating the above four points are evaluated using a real image database.