There is provided an image processing device including circuitry configured to predict, from an image of a first layer, an image of a second layer having a different pixel attribute from the first layer, by using a lookup table in which a combination of a luma component, a first chroma component, and a second chroma component of the first layer is mapped to a predicted pixel value of the second layer corresponding to the combination, wherein a predicted value of a luma component of the second layer is acquired by using a first lookup table, and a predicted value of a first chroma component of the second layer is acquired by using a second lookup table with granularity different from granularity of the first lookup table.