home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!dtix!darwin.sura.net!zaphod.mps.ohio-state.edu!cs.utexas.edu!sun-barr!sh.wide!ascwide!ascgw!gctgw!dave
- From: dave@gctech.co.jp (Dave)
- Newsgroups: comp.compression
- Subject: Re: Review: "JPEG Still Image Data Compression Standard"
- Message-ID: <DAVE.92Nov11152859@gctgw.gctech.co.jp>
- Date: 11 Nov 92 06:28:59 GMT
- References: <BxAxCE.B9C.2@cs.cmu.edu> <1992Nov9.011027.29183@cs.wayne.edu>
- Sender: news@gctech.co.jp
- Organization: GC Technology
- Lines: 44
- In-Reply-To: bhh@loki.cs.wayne.edu's message of 9 Nov 92 01:10:27 GMT
-
- In article <1992Nov9.011027.29183@cs.wayne.edu> bhh@loki.cs.wayne.edu (Bhaskar Holur) writes:
- Hi, I am doing a project on video compression using ...
- the MPEG algorithm....
-
- 1. I am looking for an image sequence (video sequence preferably)
- 512x512 (or 256x256) gray-scale, 8 bit/pel quantized, suitable
- for use on Sun Sparc 4 station terminal.
-
- Actually, if anyone knows or has a good test sequence for the
- simulation of MPEG, it would be best.
-
- If you're simulating MPEG, you may want to use one of the standard CCIR601
- formatted test pictures. They are YUV with 720 pels per line by 488
- lines. "Flower Garden" is probably the most often used sequence. Mail me
- if you want to FTP it.
-
-
- 2. Although I got quite some info from various tech. papers,
- especially from Proc., SPIE about MPEG and various motion-compensated
- prediction/interpolation encoder development and various motion-
- estimation techniques, and an overview on MPEG as well, I need more
- info for the exact implementation details for the encoder. (esp.
- for software simulation)
-
- The exact implementation details of the coder were purposefully left out
- of the standard. You can implement the coder any way you want as long as
- the output conforms to the definition of an MPEG bit stream. The aim is
- to get companies who think they can make a good coder to cooperate in the
- standardization process. There is some public domain software available,
- or at least semi-public. I have heard of some at Columbia University, but
- I don't know where yet.
-
-
- I did try to ftp into isy.liu.se for some image sequences. I was
- not able to convert from the .imf format. When I went into the
- vi editor to strip the header from the .imf file, it said line
- too long!???
-
- Your first mistake is using the vi editor. Learn Emacs. Your second
- mistake (though some may disagree) is using a text editor to modify an
- image file. I've never used the .imf format, but I imagine it is not
- composed of only ASCII characters.
-
- Dave
-