Context Coding of Depth Map Images Under the Piecewise-Constant Image Model Representation

Abstract

Abstract-This paper introduces an efficient method for lossless compression of depth map images, using the representation of a depth image in terms of three entities: 1) the crack-edges; 2) the constant depth regions enclosed by them; and 3) the depth value over each region. The starting representation is identical with that used in a very efficient coder for palette images, the piecewise-constant image model coding, but the techniques used for coding the elements of the representation are more advanced and especially suitable for the type of redundancy present in depth images. Initially, the vertical and horizontal crack-edges separating the constant depth regions are transmitted by 2D context coding using optimally pruned context trees. Both the encoder and decoder can reconstruct the regions of constant depth from the transmitted crack-edge image. The depth value in a given region is encoded using the depth values of the neighboring regions already encoded, exploiting the natural smoothness of the depth variation, and the mutual exclusiveness of the values in neighboring regions. The encoding method is suitable for lossless compression of depth images, obtaining compression of about 10-65 times, and additionally can be used as the entropy coding stage for lossy depth compression.

Patch-Based Conditional Context Coding of Stereo Disparity Images

“…A patch is a set of pixels forming a connected region (in 4-connectivity) having the same disparity value for all pixels in the region (see more formal definitions in [7]). The key for efficient coding of the image is notticing in Fig.…”

“…Incorrect warping becomes equivalent to non-existent warping. The most efficient way to encode the contours of the patches for a disparity image was introduced in CERV [7], by considering two matrices of crack-edges, and , which are marking the places where neighbor pixels in are different. In the horizontal crack-edge binary image, indicates that and in the vertical crack-edge binary image, indicates that .…”

“…In Stage of the currently proposed algorithm, encoding of and is done in interleaved manner, similar to CERV, by encoding one row from , and then one row from , which was shown to be the key for efficient coding in CERV [7]. Additionally the encoding of and is done conditionally on the crack-edge images , constructed from the warped image .…”

“…2(f) shows the image of flags . The contexts used for encoding a vertical and a horizontal crack-edge are different, similar to [7]. In Fig.…”

“…The lossless compression of a single disparity image or of a depth map has received many solutions: bit-plane decomposition was combined with binary shape coding in [2]; an "intra" method based on bitplane decomposition was proposed in [3]; encoding the contours of large regions, followed by predictive coding inside each region was used in [4]; modifications of the lossless modes of existing standards were proposed in [5] [6]. The "intra" method that is used here for encoding the right image is CERV [7], which serves also as the starting point for the newly proposed conditional coding. We use most of notations and part of the encoding strategy from CERV.…”