logo资料库

hevc test condition.pdf

第1页 / 共4页
第2页 / 共4页
第3页 / 共4页
第4页 / 共4页
资料共4页,全文预览结束
Joint Collaborative Team on Video Coding (JCT-VC) of ITU-T SG16 WP3 and ISO/IEC JTC1/SC29/WG11 12th Meeting: Geneva, CH 14 – 23 January 2013 Document: JCTVC-L1100 WG11 Number: m28412 Title: Status: Common test conditions and software reference configurations Output document Purpose: Information Author(s) or Contact(s): Frank Bossen Source: JCT-VC Abstract Tel: Email: +1 650 496 4742 bossen@docomoinnovations.com _____________________________ This document defines common test conditions and software reference configurations to be used in the context of core experiments (CE) conducted after the 12th JCT-VC meeting. These common test conditions are also recommended for use in technical contributions to the 13th and following JCT-VC meetings, if applicable. 1 Introduction Common test conditions are desirable to conduct experiments in a well-defined environment and ease the comparison of the outcome of experiments. This document defines 8 test conditions, reflecting combinations of 8 and 10 bit coding, and of intra-only, random-access, and low-delay settings:   Intra, main Intra, high efficiency, 10 bit  Random access, main  Random access, high efficiency, 10 bit  Low delay, main  Low delay, high efficiency, 10 bit  Low delay, main, P slices only (optional)  Low delay, high efficiency, P slices only, 10 bit (optional) A subset of these test conditions might be used for a particular experiment. For example, when testing an intra coding tool, only intra configurations might be used. Version 10.0 of the common software is expected to be used for most experiments. More recent versions are encouraged where applicable. Availability of the software will be announced on the JCT-VC email reflector by February 4, 2013. The following sections define test sequences, quantization parameter values, encoder configuration files, and compile-time options to be used. People bringing input contributions should provide a set of results as complete as possible that apply to the proposal. Results should be reported using the attached Excel sheets. Page: 1 Date Saved: 2014-05-02
2 Test sequences Table 1 defines the set of test sequences to be used for intra, random-access, and low-delay conditions. All frames (as defined by frame count in the table) shall be encoded for all sequences and test cases described below (see Section 4 for definitions of Main and Main10). Test sequences are available on ftp://hevc@ftp.tnt.uni-hannover.de/testsequences/ (please contact the JCT-VC chairs for login information). Table 1 Test sequences Class Sequence name Resolution Traffic 2560x1600 PeopleOnStreet 2560x1600 Nebuta 2560x1600 SteamLocomotive 2560x1600 Kimono 1920x1080 ParkScene 1920x1080 Cactus 1920x1080 BQTerrace 1920x1080 BasketballDrive 1920x1080 RaceHorses 832x480 BQMall 832x480 PartyScene 832x480 BasketballDrill 832x480 RaceHorses 416x240 BQSquare 416x240 BlowingBubbles 416x240 BasketballPass 416x240 A A A A B B B B B C C C C D D D D Frame count 150 150 300 300 240 240 500 600 500 300 600 500 500 300 600 500 500 Frame rate Bit depth Intra Random access Low- delay 30fps 30fps 60fps 60fps 24fps 24fps 50fps 60fps 50fps 30fps 60fps 50fps 50fps 30fps 60fps 50fps 50fps 8 8 10 10 8 8 8 8 8 8 8 8 8 8 8 8 8 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Page: 2 Date Saved: 2014-05-02
E E E F F F F FourPeople 1280x720 Johnny 1280x720 KristenAndSara 1280x720 BaskeballDrillText 832x480 ChinaSpeed 1024x768 SlideEditing 1280x720 SlideShow 1280x720 600 600 600 500 500 300 500 60fps 60fps 60fps 50fps 30fps 30fps 20fps 8 8 8 8 8 8 8 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Main/ Main10 Note: When the encoder operates in 8-bit mode (InternalBitDepth=8) for a 10-bit source, each 10-bit source sample x is converted prior to encoding to an 8-bit value (x+2) / 4 clipped to the [0,255] range. Similarly when the encoder operates in 10-bit mode (InternalBitDepth=10) for an 8-bit source, each 8-bit source sample x is converted prior to encoding to a 10-bit value 4*x. This behavior is built into the reference encoder and no external conversion program is required. 3 Quantization parameter values For each video sequence four quantization parameter values are to be used: 22, 27, 32 and 37. These values define the QP values used for the I-frames in a sequence (configuration files further define QP values used for other frames). 4 Configuration files The following sections define encoder configuration files to be used for each test case. Parameters to be changed for each test point are:  InputFile to reflect the location of the source video sequence on the test system  FrameRate to reflect the frame rate of a given sequence as per Table 1  SourceWidth to reflect the width of the source video sequence  SourceHeight to reflect the height of the source video sequence  FramesToBeEncoded to reflect the frame count of a given sequence as per Table 1  IntraPeriod to reflect the intra refresh period in the random access test cases. The intra refresh period is dependent on the frame rate of the source: a value 16 shall be used for sequences with a frame rate equal to 20fps, 24 for 24fps, 32 for 30fps, 48 for 50fps, and 64 for 60fps.  QP to reflect the quantization parameter values defined in section 3.  InputBitDepth to reflect the bit depth of a given sequence as per Table 1 These configuration files are provided in the cfg/ folder of version 10.0 of the common software package (available at https://hevc.hhi.fraunhofer.de/svn/svn_HEVCSoftware/tags/HM-10.0). There are 8 configurations provided as follows:  “All Intra – Main” (AI-Main): encoder_intra_main.cfg Page: 3 Date Saved: 2014-05-02
       “Random access – Main” (RA-Main): encoder_randomaccess_main.cfg “Low-delay B – Main” (LB-Main): encoder_lowdelay_main.cfg “Low-delay P – Main” (LP-Main, optional): encoder_lowdelay_P_main.cfg “All Intra – High efficiency” (AI- Main10): encoder_intra_main10.cfg “Random access – High efficiency” (RA- Main10): encoder_randomaccess_main10.cfg “Low-delay B – High efficiency” (LB- Main10): encoder_lowdelay_main10.cfg “Low-delay P – High efficiency” (LP- Main10, optional): encoder_lowdelay_P_main10.cfg Sequence-specific parameters are to be found in the cfg/per-sequence/ folder. 5 Compile-time settings Compile-time settings are defined mostly in the TypeDef.h file located in the source/Lib/TLibCommon folder of the common software. The default settings provided in the source code should be used. Page: 4 Date Saved: 2014-05-02
分享到:
收藏