IT-LIST Digest 21Topics covered in this issue include:  1) Re: IT 1.23	by  2) RE: measuring lesion area on leaves	by Don Wilcox   3) 	by Przemko Tylzanowski   4) Re: Separation of objects	by "Henrik Guldberg Pedersen"   5) RE: NT/ImageTool running slowly	by Don Wilcox   6) Unzip of IT.ZIP problem - incorrect checksum	by (Kevin R. DuPre')  7) Re: IT 1.23	by Jeff Eriksen   8) Debugging of Plugins	by  9) ImageTool Source Code	by "S. Brent Dove"  10) many questions from new user	by 11) Scanning	by "Zulmar A. ACCIOLI DE VASCONCELLOS"  12) Re: Scanning	by "David F. Bird" ----------------------------------------------------------------------Date: Fri, 10 May 1996 15:54:38 -0400From: SMooreIP@aol.comTo: it-list@sparky.uthscsa.eduCc: IPAdams@aol.comSubject: Re: IT 1.23Message-ID: <>In a message dated 96-05-10 01:37:19 EDT, (JeffEriksen) writes:>Sounds like a job for watershed.  It can separate particles much better than>simple dilation/erosion.  I vote to include watershed analysis in a future>version.  I can provide references.>>-Jeff>>Jeff,I'd be interested in learning about where I can obtain watershed.  Thanks.  Steve MooreCenter for Image Processing in Education, Tucson, AZ------------------------------Date: Thu, 9 May 1996 07:15:43 -0700From: Don Wilcox To: 'ImageTool List Server' Subject: RE: measuring lesion area on leavesMessage-ID: <>----------From:  Gregory Forbes[]Sent:  Friday, May 03, 1996 6:47 PMTo:  wilcox@xroads.comSubject:  measuring lesion area on leavesWe would like to use IT to measure leaf green area and diseased area.  =Wenow have some color images digitized and have measured both the green =anddiseased areas by tracing the perimeters, but we would like to know how =todo this by separating colors or gray scales.  The images we have =digitizedare as follows: width 640, height 480 and BPP =3D 8.  I would appreciate =anysuggestions on the best way to go about this.ImageTool has only rudimentary color capabilities right now.  You may =want to try the following:1.  Separate the color image into its components using the color =separation plug-in.2.  Apply the linear.pal color table to the individual images.  Probably =the green areas will show up best in the green sep, and the diseased =areas in the red sep.3.  Use manual thesholding in the find objects command.  You may also =want to use manual object selection (this is selected using the settings =command).If this works, you may be able to automate the steps using the macro =facility.Don------------------------------Date: Sat, 11 May 1996 11:00:14 +0100 (WDT)From: Przemko Tylzanowski To: it-list@sparky.uthscsa.eduMessage-ID: Hi!A quick question: when I import about ANY file (even TIFFs) I always run into the same problem. What I want is to look at my gels, adjust brightness, contrast etc. Whenever I try to do that, it does not work The contrast and brightness are greyed out. Once I copy it however, and then work with the copy the problem is gone. Would it be possible to make this process more aoutomatic?Przemko------------------------------Date: Tue, 14 May 1996 11:40:06 +0000From: "Henrik Guldberg Pedersen" To: it-list@sparky.uthscsa.eduSubject: Re: Separation of objectsMessage-ID: <>Hi,> > Recently several users have indicated problems associated with the > separation of joined objects for object anlysis.  Under some > circumstances traditional dilation/erosion/open/close method do not work. >  Some have suggested using manual pixel modification to separate object.> > It has also been suggested to use a watershed segmentation> > >Sounds like a job for watershed.  It can separate particles much better than> >simple dilation/erosion.  I vote to include watershed analysis in a future> >version.  I can provide references.> > Can anyone suggest a place to obtain specific implementation of the > Watershed algorithm either a place to get actual source code or a > detailed reference on the algorithm so we can include it in the next > release of ImageTool.> > Thanks for the help. There is a version of NIH 1.57, in which the watershed segmentation should be implemented. This is available from Stein Roervik ( at I think it is in the Quadra/pub/ directory. Kind regards---------------------------------------------------------Henrik Guldberg PedersenInstitute ChemistryTechnical University of Denmark, bld. 204DK-2800 LyngbyPhone: + 45 45 25 22 47Fax: + 45 45 93 48 86E-Mail: hgp@kemi.dtu.dkURL: Wed, 15 May 1996 07:32:24 -0700From: Don Wilcox To: 'ImageTool List Server' Subject: RE: NT/ImageTool running slowlyMessage-ID: <>----------From:  Denise Monteleone[]Sent:  Monday, May 13, 1996 6:48 AMTo:  wilcox@xroads.comSubject:  RE: NT/ImageTool running slowlyDon,	I believe that my problem is the size of my files. I am workingwith large 8-bit and 16-bit TIFF files. For example, a 5.6mb 8-bit TIFFfile takes approx. 1 minute to open and a 13.5mb 16-bit TIFF file takesapprox. 13 minutes to open. Then once these files are open it's slow todo anything that alters the image. Do you think more ram would solve =thisproblem? If so, are there some guide lines for how much ram is needed =forcertain size images?Most definitely memory will help.  ImageTool does not make any efforts =to speed up the display of very large images, if the image does not fit =in available RAM, then ImageTool will simply let the operating system =(NT) handle it in virtual memory.  As a general rule, you should have 8B =M of RAM for NT, plus another 2 MB for ImageTool, plus enough for the =images you wish to display.  In this case, you should have 8+2+14, or at =least 24 MB to view a single image.  If you want to view more than one =image at a time, then you would need more memory.Don------------------------------Date: Tue, 21 May 96 11:28:00 EDTFrom: (Kevin R. DuPre')To: (Image Tool Windows NT)Subject: Unzip of IT.ZIP problem - incorrect checksumMessage-ID: <>I have downloaded IT for use on NT workstation running 3.51.  In theprocess of unzipping it, I received the following messages:PKUNZIP: (W13) Warning! inconsistent local header for file INS0762.LIBPKUNZIP: (W13) Warning! file fails CRC check(it then unzips with a length of zero)Thinking perhaps this was a library not available in the freeware versionI ran setup and installed it anyway, and tried to run the application.I was able to export one of our raw binary image sequences of 500x3758-bit grayscale and it loaded 5 slices in as requested with an offsetof 20 slices from the beginning of the file. However the image looks likeit only has 3 or 4 levels of grayscale even though a histogram reveals that it has much more.The program is unstable and quits with an access privelege violation.I'm certain that the lib from the download site must be related to theseproblems, or is my suspicion that this library is not available in thefreeware version correct?  Has anybody else run into this problem?How do I fix it, or where is a download site that has an intact versionof the lib in question or IT.ZIP?Thanks in advance to all who reply,----------------------------------------------------------------------Kevin R. DuPre' 		Internet:				Personal Internet:				Whois: KRD2Motion Analysis Engineer,                    HFH.EDU Domain Admin, Security and Technical Contact,   Henry Ford Hospital		Bone and Joint Center		Motion Analysis Section,Detroit, MI 48202 USA----------------------------------------------------------------------------------------------------Date: Tue, 21 May 1996 21:26:19 -0700 (PDT)From: Jeff Eriksen To: it-list@sparky.uthscsa.eduSubject: Re: IT 1.23Message-ID: <>>Jeff,>I'd be interested in learning about where I can obtain watershed.  Thanks.  >Steve Moore>Center for Image Processing in Education, Tucson, AZ>>Steve,        If I did not get back to you before, here is some info from SteinRoervik:-Jeff**********************************************************************************Watershed segmentation is not easy to implement through a macro; you needcompiled code. I know of three available implementation of watershedsegmentation running on Macs:1) The commercial program Prism, demo on Zippy.2) Russ' new plug-in CD-ROM which contains EDM/watershed/UEP plug-ins.3) My customized version of NIH image, available through anonymous FTP   from Please use your e-mail address as password.   Select EDM/Watershed Dialog from the User-menu, on a binary image only.Stein R.**********************************************************************************If you need references, ask me again.  The above is mainly for Mac users.------------------------------Date: Thu, 23 May 96 12:01:40 BSTFrom: it-list@sparky.uthscsa.eduSubject: Debugging of PluginsMessage-ID: <9605231101.AA16186@tesla>I am currently using Borland C++ v4.52 Integrated Development Environmentto develop a plugin for ImageTool. However, the plugin fails to work inImageTool due to an ObjectWindows Exception. I have tried using the 32-bitdebugger but have failed in my attempt to load the DLL source code. It givesme the error message "No symbolic table". Having attempted to relink the DLLwith debugging information included, along with all optimisations off, I stillget the same error message!Is there something I am obviously doing wrong? Has anyone else been ableto successfully load the source code to allow debugging of their plugins?Please helpThanksJ. Glenn------------------------------Date: Thu, 23 May 1996 09:01:41 -0500 (CDT)From: "S. Brent Dove" To: IT List Subject: ImageTool Source CodeMessage-ID: <>Hello ImageTool Users,Recently we have had problems unziping the ITSDK.ZIP file which contains the source code.  Problems arise with using unzip programs that do not support long file names.  Please dowload the new WinZip application on the ftp server in the utilities subdirectory. This has been tested yesterday on the ITSDK.ZIP file.  Remember to register your copy of WinZip. Thank you again.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: Sat, 25 May 1996 20:18:15 -0500From: IT-LIST@sparky.uthscsa.eduSubject: many questions from new userMessage-ID: <>As you understand I am a new user of Image Tool ( I justdownloaded  the software about a week ago) and have a lot of questions about it. I have some problems with the threshold operation. Whenever I use it the program crashes, and I get the message "the program has performedan illegal operation and will be shut down"or "abnormal program termination" and so on. The most usual message though is "GDI failure".I use the program for quantitative analysis on brain sections so as you understand the threshold operation is very useful to me.Is it the current configuration that I use (486 PC, 8MB RAM ) or an internal problem of the program that is causing this ?Also there is a problem in opening TIFFimages created by NIH Image for Mac or even from other programs that create images in Tiff  format.Please if someone has the time to answer to my inquiries would help me very much.------------------------------Date: Sun, 26 May 1996 04:02:05 +0100From: "Zulmar A. ACCIOLI DE VASCONCELLOS" To: it-list@sparky.uthscsa.eduSubject: ScanningMessage-ID: <>Dears Fellows,I would like to make 3D images with the motoneurones of the spinal cord...I have images every 10 microns...It is possible construct the images like a CT Scan? What is the name of the program to perform it? Where can I find it?Thank you for the aid!Zulmar.Z. A. ACCIOLI, MDLaboratoire de NeurobiologieUniversite Rene DescartesParis - France------------------------------Date: Sun, 26 May 1996 12:39:59 -0400From: "David F. Bird" To: it-list@sparky.uthscsa.eduCc: info@rsinc.comSubject: Re: ScanningMessage-ID: <>At 08:53 PM 5/25/96 +0600, you wrote:>Dears Fellows,>I would like to make 3D images with the motoneurones of the spinal >cord...I have images every 10 microns...It is possible construct theimages like a CT Scan? What is the name of the program to perform it?Where can I find it?Thank you for the aid!Zulmar.Z. A. ACCIOLI, MD>Laboratoire de Neurobiologie>Universite Rene Descartes>Paris - France>Le Dr. Accioli,Je vous suggere le logiciel IDL, de Research Systems, Inc. Le programmeest disponible comme demonstration sur l'Internet a pouvez vous renseigner sur ces capabilites par courrier electroniquea Je les ai trouve tres obligeants.I suggest you try the program IDL. A demo of this program is available on theInternet at You can get further information aboutparticular applications by e-mail at I have found them to bealmost as friendly and open as the authors of ImageTool!David BirdDepartement des sciences biologiquesUniversite du Quebec a Montreal>------------------------------End of IT-LIST Digest 21