The present technology relates to an image processing apparatus, an image processing method, and a program capable of improving the image quality of decoded images of images from multiple viewpoints. A warping unit 461 warps a picture of an image of a viewpoint #1 between the image of the viewpoint #1 and an image of a viewpoint #2 different of the viewpoint #1 to generate a picture of a warped image obtained by converting the picture of the image of the viewpoint #1 into an image obtainable in the viewpoint #2. A disparity prediction unit 463 acquires, from header information, a reference index indicating a reference picture that is referred to generate a predicted image of an object block to be decoded of a picture of the image of the viewpoint #2, and selects the reference picture from among reference picture candidates including at least the picture of the warped image based on the reference index. The present technology is applicable, for example, to coding and decoding of images from multiple viewpoints.