Hi, Alex, You know, honestly, I don't know much about JPEG format too ;). However, at least simple JPEG images allow to include some alphanumeric information into the JPEG file (like copyright information, etc.). Nevertheless, I would suggest more advanced solution: 1. TV signal can be captured in usual way as it is done when capturing AVI, or Quicktime files. 2. Instead of drawing caption in the frames, your additional information can be added in headers of these files (you need to check this in AVI or Quicktime specifications, may be Rainer Johanni <rrj@xxxxxxxxxxxxxx> can help - he is the author of lavtools - utilities which we use to capture AVI files in Linux, you may also ask Gernot Ziegler <gz@xxxxxxxxxxxxxx> - he is maintainer of newer MJPEG/MPEG tools for Linux). If this formats will not satisfy your needs, you can develop your own format, it might be quite simple: you just need to store raw JPEG frames from DC10plus card along with your additional information. 3. When playing back, the JPEG frames can be sent directly to DC10+ or similar card, and your additional information can be fed into TV signal according to closed captioning (CC) specifications. In this case you will need TV monitor which has CC decoder so the captions can be seen on TV screen like captions on TV news and other programs. I think this solution is much better because it saves valuable space in TV frames (CC use only one line per field) and if you need later some kind of computer processing of your captured movie, search and identification of frame sequences will be much simple when you have your information in a raw alphanumeric format. DC10plus driver for Linux does not support CC yet but I was thinking about implementing it because ADV7176 TV encoder supports this feature and I suppose that it can be useful even for home use when somebody needs to store any additional information in his movies. To read more about closed captions, check http://frwebgate.access.gpo.gov:80/cgi-bin/get-cfr.cgi?TITLE=47&PART=15&SECTION=119&YEAR=1998&TYPE=PDF Alex Holden wrote: > On Mon, 28 Aug 2000, Serguei Miridonov wrote: > > Linux driver for Miro/Pinnacle Systems Inc. DC10plus and some DC10 cards. > > I'm currently working on a networked security camera system, and am > adapting it to work with PCs with DC10+ cards in them (previously we used > NetWinders, but they have stopped selling rack mount NetWinders with video > capture inputs). Currently we capture in uncompressed mode, draw a caption > (name of camera and time/date) on the image, then compress it in software > using libjpeg. I'd like to be able to use the ZR36060 to do the JPEG > compression, but the captions are an essential feature for us. I was > wondering if anyone has any idea how we could both draw the captions on > the image and use the hardware JPEG compression. One possibility I was > wondering about was whether it might be possible to loop the internal > video output back to the internal video input, send the captions to the > video output, and use the overlay features of the 7110A to superimpose the > captions on the captured image? > > Also, I don't know a great deal about the JPEG format, but it seems like > it might be possible to hack a JPEG file by changing the height field and > adding some more lines (containing the captions) to the bottom of the > image (or alternatively the top of the image). Does this sound feasible? > > -- > --------------- Linux- the choice of a GNU generation. -------------- > : Alex Holden (M1CJD)- Caver, Programmer, Land Rover nut, Radio Ham : > -------------------- http://www.linuxhacker.org/ -------------------- -- Serguei Miridonov CICESE, Research Center, CICESE, Optics Dept. Ensenada B.C., Mexico PO Box 434944 E-mail: mirsev@xxxxxxxxx San Diego, CA 92143-4944 FAX: +52 (6) 1750553 U.S.A.
