IT-LIST Digest 32Topics covered in this issue include:  1) Re: ImageTool Developers	by Robert Smith   2) RE: ImageTool Developers	by Don Wilcox   3) Borland 4.5 - 5.0	by davenport issac   4) ImageTool stacks and .its format	by cecil@delphi.dseg.ti.com (cecil coale)  5) Re: Borland 4.5 - 5.0	by "S. Brent Dove"   6) IT bug?	by "Fran\gois A. Richard"   7) downloading problems	by Halldor Gudmundsson   8) Re: IT bug?	by "S. Brent Dove"   9) Create area of interest based on x,y coordinates	by csu@sprynet.com 10) Re: IT bug?	by "Fran\gois A. Richard"  11) RE: IT bug?	by Don Wilcox  12) RE: TIFF LZW Compression	by Don Wilcox ----------------------------------------------------------------------Date: Wed, 10 Jul 1996 13:23:14 -0400From: Robert Smith To: it-list@sparky.uthscsa.eduSubject: Re: ImageTool DevelopersMessage-ID: <1.5.4.32.19960710172314.00698634@curtech.com>At 08:12 AM 7/10/96 +0600, you wrote:>Answering to Bob A. Smith.>If I assume that the color space is 3D, ie a cube with [0,0,0] being >black and [255.255.255] being pure white, then any 'color' { I do not >know the proper name, it is tint, hue...} is represented by a ray >from [0,0,0] to a boundary point where one of the coordinates is >255. Along that line we go from 0% intensity [black] to 100% at >the boundary. A 'distance' between 2 colors can then be defined as >the angle between the 2 vectors. A region of similarity to a given >point [r,g,b] can be an ellipse centered at [r,g,b] with major >axis on the ray through [r.g.b]. I am definitely not an expert in >colors, only a poor bio-statistician, so I was thinking on the lines >of defining distance with an ellipse, ie a quadratic in [r,g,b].>You said it has already being proven that it does not work?>Consider on the other hand, that I do not want to measure color as an >artist would define it, but I want to measure 'colorimetrically' the >concentration of a given dye, my immuno stain. The color of the dye >is modified by interfering tissue and other components of the slide >and the intensity of the color is proportional to concentration. For >this application I am not concerned with concentration, just >presence, correcting from blurring by the extra stuff.>Thanks for your comments, and if anybody else wants to comment please >feel free...>.>R. Heberto Ghezzo  Ph.D.AHA!   If you confine yourself to "hue" (whose technical meaning is roughly whatmost perple mean by "color" -- red/green/yellow) and omit brightness andsaturation (the other two dimensions commonly used to make a 3d space withhue), then you DO have a 1-d entity and you could indeed order the colors.There's a slight ambiguity in that the hue dimension is circular (i.e. fromred thru yellow, green, purple, and back to red) so given any 2 colors,there are 2 distances between them, depending on which way you traverse thehue circle.  But I don't think this is insuperable at all.  I suspect thatlooking for a distinct hue (typical of the dye) would work if the dye isquite a different color from the rest of the image.  (Many biological tracerdyes are green, because there isn't much green in the animal body.)  Still,there is no general answer to the problem of identifying a dye, because thenative colors of the biological object will always modify the hue of the dyea bit.  A common way of identifying a dye is to use one with a very sharpabsorbtion line or one which fluoresces (sp?) at a well-defined spectralline.  Since there are no lines (absorbtive or emissive) in the colors ofthe body, one can look for these with suitable filters, etc and pretty muchfilter out everything else.  I don't know if your dye has such a property,but for really accurate studies something of the sort is usually necessary.I could possibly be more help if I knew the spectral properties of your dye.Good luck.Bob             .  Robert A. Smith, Ph.D.  _____    .    Vision Systems' Analyst |     |<.      Current Technology, Inc. |_____|   .    (603) 868-2270     ^       .  ras@curtech.com    / \   /   \------------------------------Date: Thu, 11 Jul 1996 09:47:07 -0700From: Don Wilcox To: "'ImageTool List Server'" Subject: RE: ImageTool DevelopersMessage-ID: <01BB6F0F.4AD842A0@slip64.xroads.com>Hi Brent,At this moment I have the following problem:- input of my DLL is the rectangular AOI (area of interest) of the =analyzed  image. How can I get the DIB representing bitmap of selected area  (similarly as in your function =3D84SaveSelectionAs" but not writing =DIB to  the file). Or how can I use your functions to select AOI inside my =DLL.- I expect that you fill ImageClientBase and on some message is called  ImageClientBase::CmFileSaveSelectionAs. Where and when is called =Construc=3Dtor?If you look in the file iticbase.cpp, at the very bottom is the code =called to extract the rectangular AOI from the DIB.  Basically, the code =is:	RWDib *newDIB =3D new RWDib(*imageDib, left, top, width, height).This routine is provided by UTILib.  Note that the actual code does some =additional work, such as inverting the y coordinate ofthe top of the =rectangle, since DIBs run from bottom to top, and linearizing the DIBs =color table.  If this is not enough information, send me email.Don------------------------------Date: Thu, 11 Jul 1996 14:28:23 -0600 (MDT)From: davenport issac To: it-list@sparky.uthscsa.eduSubject: Borland 4.5 - 5.0Message-ID: <199607112028.OAA21213@stripe.Colorado.EDU>Someone posted a question once asking if a version of the IT SDK would beavailable that was compatible with version 5.0 or Borland.  Does anyoneknow what the incompaitbility is.  Are the 4.5 libraries no longer fullysupported?  I am thinking of buying 4.5 for this project but I wouldrather get 5.0 if it is compatible.isaac.------------------------------Date: Fri, 12 Jul 96 11:04:57 CDTFrom: cecil@delphi.dseg.ti.com (cecil coale)To: it-list@sparky.uthscsa.eduCc: cecil@delphi.dseg.ti.comSubject: ImageTool stacks and .its formatMessage-ID: <9607121604.AA03775@delphi.dseg.ti.com>I recently downloaded ImageTool from the FTP site. I want to use it to displayimage sequences which were created by a TI thermal imager. Each image is240 rows x 320 columns x 2 bytes per pixel, at a 30 Hz frame rate. These imagesare stored row wise in a binary data file. I usually view them with Matlab.I read the documentation with ImageTool and found a general description of the .its format for a stack. This sounds like what I need. I would like to somehow define my image format in the stack and then define the 200 or so images to be shown as text path names to the individual images with a text editor.  The documentation says this can be done, but supplies no example. Does anyone know how to do this?Thanks,Cecil CoaleTexas Instruments, Incms 84546600 Chase Oaks BlvdPlano, Texas 75023(214) 575-7973------------------------------Date: Fri, 12 Jul 1996 13:00:45 -0500 (CDT)From: "S. Brent Dove" To: IT List Subject: Re: Borland 4.5 - 5.0Message-ID: <01I6ZIN3PP1I00CA37@uthscsa.edu>Recently on ITLIST>Someone posted a question once asking if a version of the IT SDK would be>available that was compatible with version 5.0 or Borland.  Does anyone>know what the incompaitbility is.  Are the 4.5 libraries no longer fully>supported?  I am thinking of buying 4.5 for this project but I would>rather get 5.0 if it is compatible.The current ImageTool SDK is for Borland C++ 4.52.  As you may or may not know Borland's C++ 5.0 compiler required some modification of existing code to compile.  We currently have a recompiled ImageTool under 5.0 and are presently testing the software.  We hope to have this testing completed by next week.  We will continue to have online version 1.23 and its SDK which is for Borland 4.52 and the next version 1.25 which compiles with Borland 5.0.  We will continue all future developments under the new Borland 5.0 C++ development environment.  We intend to make a few additions in the version 1.25 and get it out as soon as is possible.  We are trying to test the software as much as is possible, but we will rely on our users to test and give us feedback on any bugs that may be found.  So long for now.S. Brent Dove                            Voice: (210) 567-3333Diagnostic Sciences                      Fax:   (210) 567-3334University of Texas                      Email: dove@uthscsa.eduHealth Science Center                    Web:   ddsdx.uthscsa.eduSan Antonio, TX    USA                   ftp:   maxrad6.uthscsa.edu------------------------------Date: Fri, 12 Jul 1996 14:23:59 -0400From: "Fran\gois A. Richard" To: it-list@sparky.uthscsa.eduSubject: IT bug?Message-ID: <31E6983E.65AD@geol.uottawa.ca>S. Brent Dove wrote:[stuff deleted...]> We are trying to test the software as much as is possible, but> we will rely on our users to test and give us feedback on any bugs Here's one for you: each time I try opening a new stack (Stacks|New), ITcrashes, giving me two error messages ("IT caused an invalid page faultin module IT.EXE at 0137:00455ce9" (repeated 4 times), and "IT caused anexception c0000026H in module KERNEL32.DLL at 0137:bffb87c0").This may not be a bug at all, but everything else on my PC still worksas well as it did before. I really hope someone can figure out what theproblem is, because this feature (stacks) is quite useful to me.Thanks for any hints/suggestions/(solutions?)...-- Francois A. Richard            Email: Richard@geol.uottawa.caUniv. of Ottawa (Geology)      Tel:   (613) 562-5800, ext. 6850Ottawa (CANADA) K1N 6N5        Fax:   (613) 562-5192=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= http://www.geol.uottawa.ca/geo/lists/grads/richard/richard.htm------------------------------Date: Mon, 15 Jul 1996 12:03:04 GMTFrom: Halldor Gudmundsson To: IT-LIST@sparky.uthscsa.eduSubject: downloading problemsMessage-ID: <199607151203.AA164332184@iti.is>Dear colleagues,I am unable to download the DLL.ZIP file for Image Tool version 1.23.  I candownload the IT.ZIP file easily but the download for the DLL.ZIP file neverfinishes, i.e. I always get something less than the 997998 byte file size.I have used both the direct download using Netscape 2.01 and using an ftpprogram (Ws_ftp) to get the file from maxrad6.uthscsa.edu.  This hashappened about 4 times.  Are there other sites which I might attempt toaccess or can I get the file by other means???  I have usually attempted todownload the file between the hours of 3-6 AM Texas time (assuming that I amdownloading from there).Many thanks,Halldor GudmundssonHalldor Gudmundsson     |  Halldor.Gudmundsson@iti.isProject manager |Technological Institute of Iceland	tel: +354 - 587 - 7000	Keldnaholti, IS-112 Reykjavik	fax: +354 - 587 -  7409Iceland------------------------------Date: Fri, 12 Jul 1996 16:01:05 -0500 (CDT)From: "S. Brent Dove" To: IT List Subject: Re: IT bug?Message-ID: <01I6ZOXP9AM000CA37@uthscsa.edu>Recently on ITLISTRichard stated>Here's one for you: each time I try opening a new stack (Stacks|New), IT>crashes, giving me two error messages ("IT caused an invalid page fault>in module IT.EXE at 0137:00455ce9" (repeated 4 times), and "IT caused an>exception c0000026H in module KERNEL32.DLL at 0137:bffb87c0").I assume you are running the latest version 1.23 of ImageTool.  I have tested this feature on six of our test machines (486s, pentiums, laptops) under Windows95 with no problems. Have any other users experienced this problem.  Remember that images within a stack must be the same size image matrix (ie. height and width in pixels) as well as the same bit depth.  You can test this by selecting New from the Stacks menu and open images of the same size and bit depth (p.s. this could be the same image over and over). As each image is opened it will be added to the stack.  If a few users will try this and send some info just a note that you tried it and the result, it would be greatly appreciated.  If we can not reproduce the bug we cannot kill it.So long for now.S. Brent Dove                            Voice: (210) 567-3333Diagnostic Sciences                      Fax:   (210) 567-3334University of Texas                      Email: dove@uthscsa.eduHealth Science Center                    Web:   ddsdx.uthscsa.eduSan Antonio, TX    USA                   ftp:   maxrad6.uthscsa.edu------------------------------Date: Mon, 15 Jul 1996 09:00:15 -0700From: csu@sprynet.comTo: IT-LIST@sparky.uthscsa.eduSubject: Create area of interest based on x,y coordinatesMessage-ID: <199607151600.JAA23317@m1.sprynet.com>We would like to transfer a Region of Interest from a CT scan image to a Gamma Camera Image.  We have both of these images as tiff files.  We have the x,y coordinates (10 points)of the Region of Interest from the CT scan Image in pixels that have been adjusted to the size of the Gamma Camera image. Is there a way with ImageTool to create a Region of Interest on an image based on these coordinates ? ------------------------------Date: Mon, 15 Jul 1996 12:17:34 -0400From: "Fran\gois A. Richard" To: it-list@sparky.uthscsa.eduSubject: Re: IT bug?Message-ID: <31EA6F1E.4563@geol.uottawa.ca>S. Brent Dove wrote:> > Recently on ITLIST> Richard stated> > >Here's one for you: each time I try opening a new stack (Stacks|New), IT> >crashes, giving me two error messages ("IT caused an invalid page fault> >in module IT.EXE at 0137:00455ce9" (repeated 4 times), and "IT caused an> >exception c0000026H in module KERNEL32.DLL at 0137:bffb87c0").> > I assume you are running the latest version 1.23 of ImageTool.  I have> tested this feature on six of our test machines (486s, pentiums, laptops)> under Windows95 with no problems. Have any other users experienced this> problem.  Remember that images within a stack must be the same size image> matrix (ie. height and width in pixels) as well as the same bit depth.> You can test this by selecting New from the Stacks menu and open images> of the same size and bit depth (p.s. this could be the same image over> and over). As each image is opened it will be added to the stack.  If a> few users will try this and send some info just a note that you tried it> and the result, it would be greatly appreciated.  If we can not reproduce> the bug we cannot kill it.> > So long for now.> > S. Brent Dove                            Voice: (210) 567-3333> Diagnostic Sciences                      Fax:   (210) 567-3334> University of Texas                      Email: dove@uthscsa.edu> Health Science Center                    Web:   ddsdx.uthscsa.edu> San Antonio, TX    USA                   ftp:   maxrad6.uthscsa.eduJust a quick note regarding this issue: I can't even get to the point ofopening images. The program crashes as soon as I hit the Stacks|New menuitem.-- Francois A. Richard            Email: Richard@geol.uottawa.caUniv. of Ottawa (Geology)      Tel:   (613) 562-5800, ext. 6850Ottawa (CANADA) K1N 6N5        Fax:   (613) 562-5192=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= http://www.geol.uottawa.ca/geo/lists/grads/richard/richard.htm------------------------------Date: Tue, 16 Jul 1996 10:20:42 -0700From: Don Wilcox To: "'ImageTool List Server'" Subject: RE: IT bug?Message-ID: <01BB7301.AF126340@slip11.xroads.com>I have managed to discover the source of this bug, and it will be fixed in the 1.25 release of the software.Don------------------------------Date: Tue, 16 Jul 1996 10:22:43 -0700From: Don Wilcox To: "'ImageTool List Server'" Subject: RE: TIFF LZW CompressionMessage-ID: <01BB7301.B59F5420@slip11.xroads.com>----------From:  Denise Monteleone[SMTP:dcm@image.bio.bnl.gov]Sent:  Friday, July 12, 1996 12:58 PMTo:  wilcox@xroads.comSubject:  TIFF LZW CompressionDon,	I need to open TIFF files that are LZW compressed and I have notbeen able to. Is this a problem with the software or I'm I doing =something=20wrong?The problem is that Unisys now enforces a patent on the LZW algorithm =used in Tiff images, and we have not succeeded in getting any =information from them about licensing the technology.  Until this issue =is resolved, ImageTool will not be able to open either Tiff LZW or Gif =images.Don------------------------------End of IT-LIST Digest 32