Removing Depth Map Coding Distortion by Using Post Filter Set

Norishige Fukushima , Tomohiko Inoue, Yutaka Ishibashi

We propose a fast post filter set for removing coding distortions of depth maps. Lossy coding generates various distortions, such as mosquito and block noises, edge blurs, and over quantization. These distortions seriously deteriorate image quality of synthesized views in free viewpoint image rendering. Thus, we propose the post filter set which includes median filter, Gaussian filter, min-max blur remove filer, and binary weighted range filter to remove these noises.

In experiments, we use various codecs, such as JPEG, JPEG-LS, JPEG2000, and H.264/AVC, for depth map coding, and synthesize views with the coded depth maps. Experimental results show that our post filter set can improves every codecs performance. Improvement of PSNR is larger than the conventional post filter, and especially JPEG is large. After the post filter set, coding performance of JPEG with arithmetic coding reaches H.264 intra-coding. The computational time of the filter set, which is implemented by C++ with SIMD optimization, is within 5.2 ms at high bit rate cases, and within 15.3 ms at low bit rate cases.

The code is written in C++ with SIMD intrinsics (SSE4.1) and TBB parallel optimization, and uses OpenCV 2.45 and Intel Threading Building Blocks (Intel TBB).

Help for the app

The code encodes depth maps or disparity maps from MS Kinect or stereo matching, and then renders free viewpoint images by point cloud rendering. You can enable/disable various configurations in config.h, such as TBB and SIMD optimization, libjpeg-turbo option, and OpenNI Kinect Input.

Kinect depth map coding

  1. N. Fukushima, T. Inoue, Y. Ishibashi, "Removing Depth Map Coding Distortion by Using Post Filter Set," Proc. IEEE International Conference on Multimedia and Expo (ICME 2013), July 2013.
    author  = {N. Fukushima and T. Inoue and Y. Ishibashi},
    title   = {Removing depth map coding distortion by using post filter set},
    booktitle = {Proc. IEEE International Conference on Multimedia and Expo (ICME)},
    year    = {2013},