|
3 May 1999
Testimonials and other Tidbits (V6.10)The following are some of the complimentary written comments and related interesting items received at MicroImages during the last quarter exactly as provided except for the [edit] alterations in [brackets] to keep them anonymous where necessary. Many more comments are received by MicroImages by voice but cannot be reproduced here verbatim as quotes. Please note that these quotations are not edited from their original form in spelling, grammar, punctuation, and so on. From GeoWorld Magazine What follows are excerpts from a long, detailed article discussing 6 image processing systems. The full article appears in the May 1999 issue of GeoWorld, pages 36 to 46 and is titled "Image Processing Software: System Selection Depends on User Needs", by W. Fredrick Limp, Director, Center for Advanced Spatial Technologies, 12 Ozark Hall, University of Arkansas, Fayetteville, AR 72701. The image processing products discussed in this article are TNTmips 6.0, ENVI 3.1, ERDAS IMAGINE 8.3.1, ER Mapper 6.0, Image Analyst 7.0, and PCI 6.3. Extracts of the article which deal specifically with TNTmips and TNTlite are quoted here with comments. Major sections of this article have been omitted. Copyright prevents the reproduction of the whole article here, so please obtain a copy of it for specific details and commentary, pros and con, on the other competing products. If reprints of this article are purchased by MicroImages, they will be included in our next mailing. Fred Limp, the author of this review, is Director of the Center for Advanced Spatial Technologies at the University of Arkansas (CAST). A visit to the CAST web site (http://www.cast.uark.edu/cast/info/labs.mgsl.equip.html) will show that they report more than 50 workstations, peripherals, and software systems are available, running primarily Intergraph’s MGE and image processing software. The following additional software is listed as in active use at CAST in varying numbers: GRASS 4.1 Arc/Info rev 7.0 MapInfo 2.0 PCI 5.2 ERDAS Imagine 8.1 IMPORTANT NOTE: CAST does not use TNTmips, TNTlite, any of the TNT products, or ENVI! As a university site, CAST would have access to ENVI and ERDAS products at almost no initial cost. "Software evaluations often involve creating a list of features to which each package is compared, and the software with the longest list wins. But that approach doesn’t consider how the real world works. If the same logic was applied to automobiles, everyone would drive James Bond’s latest Aston Martin, because it has the most cool gizmos. In the real world, however, some people want pick-up trucks, some need six passenger sedans, and a few (a very few) need Aston Martins with a Bond girl in the passenger seat." "The giant review list system also leads to bloated software–as developers respond to ‘gotcha lists’ and their often unrealistic demands. Outside the clean pages of a magazine review, software don’t fall into neat orderings from best to worst. There can be turkeys as well as stars, but software are designed with a particular target in mind, and they probably do that ‘thing’ really well. This is particularly true in the arena of image processing software." "Image processing has moved from being a skill practiced by a few specialists to something that can potentially be found on any user’s desktop. There are many different types of sensor systems, many different domains in which image processing is used and many different levels of use. Other key factors are the size and complexity of the data that will be processed, and the requirements of the image processing system to interoperate with other systems. The real problem today isn’t to identify the ‘best’ image processing system, but to find the best fit between users’ needs and available systems, which is the goal of this review." Multiple Data Format Interoperability "The ability to handle needed data is essential. It may be that only one format is critical, or many different datasets may be needed. Users may need to access raster data, vector data or produce vector and raster output in different formats. With several of the packages, it’s possible for a competent programmer to develop routines that import or export data. But this review only considers built-in capabilities." "All the software in this review import or read the standard remote sensing data types. ENVI, in particular, expands this to cover several hyperspectral sources (discussed in more detail on page 41). However, PCI, ER Mapper and TNTmips have substantial raster and vector capabilities." "For example, in addition to options listed in the tables on pages 40, 42 and 43, TNTmips also can import vector formats such as MIF/MID, MOSS, SIF, BNA, TYDAC and VPF, and raster formats such as Laser-Scan, GRASS, EPPL7, SPANS and others–a total of 58 formats, according to the manual. [The tables and this paragraph do not mention the important MapInfo TAB format which TNTmips imports]. PCI, through its GeoGateway module, clearly has focused on interoperability, going beyond import/export to read and write capabilities for several different data formats in their native structures. This ability can be important, because multiple software can use the same data without having to create different versions through an import." "Bottom line advantage for multiple data format interoperability: PCI and TNTmips, with a nod toward ER Mapper." [The author has overlooked that a variety of formats can be linked to directly and used by TNTmips processes without the need for importing. Also, our editing procedures directly operate on popular vector formats such as ESRI’s native coverage and shapefiles and MapInfo’s native TAB files. TNTmips 6.1, issued after this review, imports several popular types of hyperspectral images such as AVIRIS and even imports hyperspectral images from the ENVI format, but ENVI does not import from our format.] Visual Display and Enhancement "One of the more common and important image processing operations is the visual display and enhancement of images on a computer screen. Applying different filters and other methods to highlight otherwise hidden aspects of an image can play a critical role in analyses. All the packages provide solid basic tools for routine image display, enhancement and filtering. ERDAS has a nice ability to display multiple views (with different zooms, etc.) of the same dataset while changes can be reflected in another window." "Bottom line advantage for visual display and enhancement: Too close to call (possibly a slight nod to ERDAS)." [TNTmips certainly has the multiple view window and interview operability. TNTmips and TNTview provide all these features of ERDAS and many more that ERDAS does not!] Classification Methods and Procedures "Effective classification procedures are key to any image processing package that handles multispectral data. Historically, the procedures have been grouped into two classes: supervised and unsupervised. Supervised classification typically involves users defining target areas onscreen (e.g., a corn field) and ‘asking’ the software to identify spectrally ‘similar’ pixels. Unsupervised classification typically involves the automatic creation of spectral grouping that are ‘similar’ in a statistical sense. Users then assign an information class (label) to each grouping. Modern practices are expanding and blurring this distinction, but it’s a useful starting point." "All the packages have more-than-adequate basic classification capabilities. If that’s satisfactory for the job, then any package is a good choice. There’s variation in the packages, however, in how users can influence and/or control the classification process and assess the results. But it’s beyond the scope of this review to compare and contrast the various methods. The key point for prospective users is to carefully assess the types of classification methods needed, and examine each package with those needs in mind. Also, there are a range of newer techniques to be considered, including fuzzy and neural-net classifiers, K means, step-wise linear discriminate functions and others." "Bottom-line advantage for classification methods and procedures: All are adequate for basic classification work. ERDAS IMAGINE, PCI and ENVI have particularly extensive options. TNTmips provides the broadest range of classifiers in its basic package." Rectification and Geometric Registration "The association of geographic coordinates to imagery matrices is a critical process, and there are several methods that can be applied. Some processes may involve an understanding of the location properties of the sensor (aircraft or satellite), and other methods use known locations of points and compute a transformation to change the image points to the known locations. This process is particularly challenging, when imagery is produced from an aircraft, because aircraft regularly move in three dimensions. Completely different algorithms are needed to properly process this type of sensor data." "In addition, creating a seamless image from aerial sensors can be challenging for multispectral sensors, because aircraft may be flying one direction (e.g., northwest) during one pass and an opposite direction (e.g., southeast) during the adjacent pass. In this case, the sun angle is different and must be adjusted before effective area analysis can proceed. A special case of this process is associated with creating orthophotos from aerial photography, which is covered in the next section." "In general, there are three aspects to the rectification process: definition and association of ground control points (GCPs), transformation of image locations to ground locations, and resampling an image to a new structure. All the packages do a more-than-adequate job of the traditional process of locating points on an image and associating them with GCPs. And all packages provide the basic transformation processes of first-, second- and third-order transformations. ER Mapper, TNTmips and ENVI in their base packages, and PCI and ERDAS IMAGINE in advanced packages, provide a triangulation transformation that divides an image into small triangles and applies different transformations to each triangle. This approach can be effective for aircraft sensor data." "Bottom-line advantage for rectification and geometric registration: No clear differentiation." Creating Orthophotos "Creating orthophotographs from aerial photography is an increasingly common task. All the vendors can create orthophotos, but they do so in quite different ways. Basic orthophoto creation is provided for in the ENVI, ER Mapper and TNTmips packages and in the advanced ERDAS IMAGINE and PCI packages. High-end, production orthophoto capabilities are available from PCI, ERDAS and Intergraph, but with substantial price tags. ER Mapper, though it doesn’t have some of the high-end orthophoto creation capabilities, has a nice capability to generate seamless mosaics from orthophotos (or similar imagery)." "Bottom-line advantage for creating orthophotos: For basic orthophoto capabilities: ENVI, ER Mapper and TNTmips. For high-end capabilities: PCI, ERDAS IMAGINE and Image Analyst, with Image Analyst the common choice for industrial users." [But the real task is the creation of DEMs. We in the United States seem to lose track of the fact that most nations do not yet have the DEMs needed as input to create orthophotos. The TNTmips process starts by creating DEMs from stereo pairs. Some competing products do not create this important product, and those that do require the very expensive add-on packages created by some other companies. The comment on mosaicking is off-hand and out of place, as it is not related to this topic. Several of the other products have mosaicking capabilities which are not covered anywhere in this article.] Hyperspectral Data Processing and Analysis "Hyperspectral data are of increasing interest within the remote sensing community. And with new sensors coming soon, hyperspectral analysis will undoubtedly grow in importance and use. Although there are various definitions, a common current hyperspectral sensor is the Airborne Visible Infrared Imaging Spectrometer (AVIRIS), which provides 224 tightly delimited spectral bands." "In some ways, hyperspectral analysis is ‘simply’ an extension of traditional multispectral analysis (a position taken by ERDAS, for example). In other ways, it introduces a different set of problems and opportunities. Because of the larger number of bands, display and analysis of hyperspectral data can be challenging. Hyperspectral analysis techniques include multidimensional plots, the ability to set wavelength ranges and save image spectra in a library, linear spectral unmixing, reflectance calibration and spectral angle mapping." "In addition, it has become possible to create reference spectra for various minerals and other materials. Unknown data can be compared to these reference spectra for identification purposes. Two common sources of reference spectra are the U.S. Geological Survey and NASA Jet Propulsion Lab spectral libraries, which are provided by ERDAS, ENVI, PCI, and TNTmips." "ENVI, TNTmips, and PCI have a substantial emphasis on hyperspectral data, ERDAS has less, and ER Mapper and Image Analyst have none. It seems clear that ENVI currently has a dominant position in this area, and it provides an extensive suite of tools to handle hyperspectral data. TNTmips also has a good suite of tools for hyperspectral analyses. In fact, the free version of TNTmips, TNTlite, can process and fully analyze AVIRIS scenes." "Bottom-line advantage for hyperspectral data processing and analysis: This area will be one of great competition, but ENVI currently leads with TNTmips second." [Of course, MicroImages thinks TNTmips is number one in this area and incorporates all the hyperspectral processes in ENVI and some more they do not provide. As noted, the price is also right as it is all free via TNTlite.] Usability, Documentation and User Interface "Documentation is a critical aspect of the usability of any software product. Good documentation will increase productivity by minimizing the time spent in incorrect processes and providing quick access to specific information needed to complete a task. Also, users often complain that a product ‘can’t do X’ when, in fact, the software can, but it’s not well documented." "The user interface also is key to a software system’s usability. A well-designed wizard system that prompts users for correct answers, or sometimes figures the answers out, lessens the need for documentation. The user interface issue, however, is more complex. Wizards can be effective for new users, but they can get in the way of experienced users or ‘power’ users. Therefore, a user interface should be considered from each perspective." "In addition, it’s not uncommon for software designers to complain that users ‘didn’t read the manual,’ and, having been involved in many software development efforts, I certainly can appreciate this point of view. There’s also often a ‘macho’ attitude toward interfaces: ‘I only need a command line prompt, vi and awk–I don’t need wimpy wizards.’ This is particularly true for software developed on UNIX platforms. As much as people like (or not like) to be ‘real’ computer geeks, it’s clear that good interface design and help systems dramatically increase user productivity, decrease training costs and increase the potential market size for any product. Image processing is a complex business, but it clearly can benefit from solid interface design." "The documentation maximalist award winner was TNTmips–a small fork lift dropped off two large boxes of information. One box contained three large, three-ring binders (with an estimated 3,000 pages of text) and three small, three-ring binders (with an estimated 900 pages of documentation). I say estimated because each section of the manuals has its own numbering and table of contents, but there’s no overall index or pagination. Users, however, can access and search the online versions." "The larger manuals serve as the user’s manual, providing specifics on each aspect of operation. The three smaller manuals explain concepts and process sequences. In addition, there are excellent color images that show operation results. The TNTmips documentation is thorough, and would be excellent if it weren’t so difficult to find things (there are no indexes). The online versions make global searching and discovery somewhat easier." "TNTmips has a wide range of GUI interfaces, but like most of the other multiplatform packages, there’s no comprehensive wizard system. Work flow steps are well outlined in the three ‘Getting Started’ manuals. In addition, TNTmips has a batch jobs capability with a substantial amount of job control." [The "three Getting Started manuals" referred to above are the "three small, three-ring binders" in which he was provided all 45 of the printed Getting Started booklets.] Value "Value is in the eye of the beholder–if a package does what users need, then it’s clearly the most valuable for the user. But value also can be looked at from some standard viewpoints. The first would be the cost to obtain basic functionality. What’s basic is arguable, but it should include the operations necessary to access and view data, perform basic classification and output results. The next level of value (intermediate) includes more advanced capabilities in each area, including modeling and operations such as image mosaicking and creating good cartographic output. The next level (advanced) would add the ability to process hyperspectral and radar products and create orthophotography. There are other ways to slice the loaf, but this seems fairly standard." "The best value in the pack is TNTlite, which includes the full range of TNTmips capabilities but limits the dataset size and output options. It’s free as a download, or it can be obtained for a modest cost. The maximum dataset size is exactly that of a single AVIRIS scene, so if users just need to perform AVIRIS analysis, then this package is the way to go." [The word modest is certainly an understatement, as the TNTlite CD cost of $10 and the kit cost of $50 seem very modest in comparison to the several "entry level" products noted below with prices ranging from $2000 and up.] "These entry level versions, however, have substantially less functionality than the next group, which includes TNTmips (prices range from $3,000 to $6,000, depending on the screen resolution options selected), ENVI at $3,300, ER Mapper at $4,300, ERDAS Advantage at $4,000, Intergraph’s Image Station at $6,000 and the PCI EASI/PACE Image Processing Kit at $4,900. At these prices, users get the complete versions of ENVI (excluding ENVI’s programming language, IDL, which is an additional $1800), ER Mapper and TNTmips. The ERDAS Image Station and PCI packages at these prices, however, don’t include advanced classification and/or radar capabilities. Therefore, users limited to this price range and need probably would prefer the other packages." "Bottom-line advantage for value: The best deal, no question, is TNTlite–it’s free (but it’s only for small data sizes). At or less than $2,500, its a ‘toss-up’ between PCI and ERDAS. The $3,000 to $6,000 range is the market’s ‘sweet spot’ with a lot of choices. TNTmips gives users a complete GIS and image processing package. In this price range, neither the ERDAS, Image Station or PCI bundles have the range of capabilities that ENVI or ER Mapper have. If users have a limited budget but need basic orthophoto and/or radar functionality, then TNTmips, ENVI or ER Mapper may be better choices, for the money, than ERDAS, PCI or Image Analyst." [We congratulate this author who can tell the difference between a real, functioning product such as TNTlite and the demonstration products of our competitors.] Overall Assessments [Each of the 6 image processing products are summarized here, please see the original article for details on competing products.] "TNTmips is in a class by itself. It’s modestly priced (from $3000 to $6000 on Microsoft Windows-based platforms, depending on resolution), runs on every type of computer, provides an amazing range of functions and is a competent raster and vector GIS. Also, the amount of documentation is astonishing, but it’s not well organized and indexed. The answer is there, but it’s hard to find. TNTmips has good interoperability with other software and data, and, if users can limit themselves to small datasets, they can run TNTlite, which is free! For those looking for a GIS and image processing solution, TNTmips may be the best single solution." [The power users of image processing software mentioned earlier by this author provide constant input to MicroImages and, we assume, to these other vendors requesting new and often obscure and exotic features. MicroImages tries to respond promptly to add as many of these features as possible and provide them to our clients through frequent upgrades. Because we do respond, we get even more requests, which helps us make better software for you. Subsequently, as the software evolves, these features have to be tucked away deeper into the user interface to maintain a simpler "look and feel" for the beginning user of IPS–LegendView, introduced in V6.10 is an example of a simplification. Rapid responses to client requests is not the policy of competing products which issue upgrades to their products on about 18 month cycles or only upgrade a certain subprocess and fix errors in others in shorter cycles.] [One of the results of this rapid and important evolution is that you have the latest tools that can be invented by MicroImages, you, or some other related product you have seen. However, frequent upgrades make it difficult to maintain documentation and self-help guides such as our Getting Started booklets. As a result, it has not been possible to provide a single table of contents in these materials based upon a single page numbering system across all documents.] [The new context HELP system being introduced with V6.10, when fully implemented, will provide an on-line "table-of-contents" which can keep up, without having continuous page numbers, with the various locations, relocations, expansion, and other constant alterations of the reference manual and to some extent the Getting Started booklets. Since most of you are using the on-line reference manual, the perfection of this new scheme will overcome the author’s criticism with an even better approach.] [The Getting Started Booklets were deliberately designed to "stand alone" as separate self-help tutorial booklets. This format was selected to encourage new users or those unfamiliar with a process to sit down and spend a few minutes to "try-it-out" by running through the examples or at least through the booklet. However, this design did not anticipate the large number of individual booklets which have now been prepared. It is now confusing to a new user where to start among these booklets. It is possible that this can be overcome by providing reference sheets or diagrams showing which booklets to cover first to get up to speed in managing and visualizing geodata and then which additional booklets should be reviewed before starting out upon a specific kind of project. A single, comprehensive index covering all the booklets, probably electronic and on-line, is also being considered.] From MicroImages clients using TNT professional products Extracted from email from Romania on 24 December 1998 "Our current MIPS key is indeed extremely busy. This situation was somehow caused by TNTlite: it allows all our students to explore and learn the functionality of TNTmips, and those continuing their studies to MSc or PhD level are asking for the professional MIPS to help them in preparing their work and dissertation. So we decided to have a second license, as the current MIPS system is busy 24/24h (interactive work during day and batch processes during the night), and the current demand is even higher than that." "We’ll have more funding next year (a larger sum), when we will discuss the upgrade of the existing key and some additional license (maybe 5 seat lab). For now we will be happy to get the second key." [currently buying a second system] "Please also note that, if you have a prospective client in Romania, you may direct him/her to our University, we will be glad to offer a free of charge MIPS demo. We have already presented and/or demonstrated TNTmips at a couple of symposia or scientific meetings, as we have a high degree of satisfaction from our MIPS system." Email from the USA on 14 January 1999 "I read the version 6.0 release materials today and thought I had better write while the ideas are still fresh. It’s 1:00 a.m. and usually hard to keep my eyes open after 9 hours of work and 3 hours of family. But even while lying in bed I was awake cover to cover while reading the materials. I realize that the written materials are as important as the software. Three parts really hit me; the grassland letter of intent, the testimonials on the university classes, and the flier ‘A True Story’. I am really starting to feel that the products are knocking down the ivory towers. The letter of intent was very personal, which I think is unusual in science. But the material was the most relevant description of misdirected focus of remote sensing and precision ag I have seen." [some personal notes omitted here] "The testimonials on the number of students interested in geospatial analysis were very rewarding. I have been wrestling with the idea of offering training classes for some time now, and after that bit of info I will try to set up extension courses to teach geospatial analysis with TNTlite. I think I will set up the courses at universities other than [a name]. The other dealer testimonials also supported my conception that Arc/Info users are not really satisfied with their products, just tolerant of poor quality and POWERLESS to do anything about it. Imagine running TNTmips completely from SML? [referring here to the use of AML in Arc/Info] We might as well do back to command line execution. And this is what ESRI users are still doing, capturing commands as macros and repeating them over and over. TNTlite is giving geospatial students power." "The flyer ‘A True Story’ completely describes the ‘joke’ being played on the federal government. I will post it on my wall. Competent integrated systems engineering is completely dependent upon thorough, detailed trade study analysis of commercially available software, even if it means delaying deliveries. Without a detailed cost/benefit review, hundreds of thousands of dollars will be wasted on expensive, limited-purpose applications." [Hundreds of millions have already been wasted! For example, in January 1999, just one US Federal Government agency, the US Bureau of Management, admitted total failure and canceled the contract after 6 years and $400 million dollars wasted for development of their ALMRS (Automated Land and Mineral Record System) for a GIS and related records management software and hardware system!] "Great release materials once again I am rejuvenated. Sorry I can’t make the workshop." Extracted from email from South Africa on 25 January 1999 "I still haven’t gone through all the new features, but one thing I thoroughly enjoy is having the cross on the second image when displaying two images, so one knows exactly where you are on both images." Extracted from email from South Africa on 1 February 1999 "I am a fairly new user of TNTmips and am thoroughly enjoying it thus far (lots and LOTS to learn!). I am looking for information regarding how TNTmips can fulfill our corporate needs as I will explain below. [A dealer] is being of great help in a number of areas (nuts and bolts) concerning this issue, but there are some other issues that I would like to ask you about in order to assist me in providing the relevant information to the powers-that-be (PTB)." [Some details skipped here] "Our corporate system (UNIX) is handled to some extent by Genamap, but as you may be aware, Genasys is going through a crisis and the possibility of switching software is now up for discussion. I would obviously favour a total switch to TNT for a variety of reasons, and have a fair deal of support for doing so." [continues with questions] Extracted from email from Italy on 10 February 1999 "I have started work with some of the tutorials and they are excellent, really first rate. I have some real data to work with (projects with aerial photos, vectorization of elevation maps, etc.) and hope to have good results quickly."
Extracted from email from Alaska on 12 February 1999 "We have been using your software since before windows and it keeps getting more useful and user friendly all the time." Extracted from email from Great Britain on 10 March 1999 [From a client with the current versions of ENVI and TNTmips] "ENVI vs. TNT. I don’t like ENVI very much, OK for experimentation but it isn’t a production system, it’s quite ‘clunky’, and doesn’t have the breadth of vector/cad/database functionality that TNT has. Also it generates a huge number of individual files (images, gcp’s, stats) that you end up struggling to manage. I have no procedure for feeding ideas to them." [The integrated Project File concept is one of the unique features of the TNT products.] Extracted from email from Great Britain on 11 March 1999 [With reference to his student who is currently using a TNTmips 5.0, circa 1995] "The student is working on a research project with me, and I have recommended that she use TNTmips because it is a complete package. More over, since she needs to learn the ropes and I am writing a book about that, she will be an ideal tester for what I produce." "Apart from that she will not be imputing to the book project, except to try out my exercises using TNTlite v. 6.0." "In many ways v.5.0 is still a viable package, certainly for what she will need to do, and in some cases such as vector editing I think that it is somewhat easier to master than the new material, though not so comprehensive." Extracted from an email exchange with France on 27 April 1999 [From MicroImages concerning questions regarding processing of CASI hyperspectral images] "It is hoped that your introduction to TNTmips has been uneventful given the Getting Started booklets and [another French client’s name] instruction? Hyperspectral image analysis is one (1) of the areas that MicroImages has focused considerable attention on over the last few software releases and the release for version 6.10 is no exception. A list of the new features added [by V6.10] to the hyperspectral process was sent to all clients in late March. You may wish to review this list to see how they can be applied toward your research." [From client] "Thank you, my introduction was good and Getting Started booklets helped me in this way." [From MicroImages] "Your interest in using CASI imagery is of particular note. This is due to the fact that the president of MicroImages, Dr. Lee Miller contacted the president of the company who makes the CASI hyperspectral device about (4) months ago. Our call was in reference to supporting the native format outputted by the CASI device as a standard import in TNTmips. At that time, we were promised information on the format as well as sample imagery. However, as of this date, nothing has arrived. Do you have any sample imagery and format documentation? If not, it might prompt some action if there was a request by a user of the CASI data for having support within TNTmips for a direct import routine." [From client] "In fact, post-processing software provided by ITRES (the company who makes the CASI hyperspectral device) reads the data on the band recorded during the flight in the native format and performs roll and geo correction. The output data provided by this software are in PCI format (.pix). So it is possible to import directly the files from CASI with the [TNT] import PCI routine." Extracted from email from Poland on 27 April 1999 [Client using a multiple user TNTmips in UNIX] "I’m sorry that you wait so long for the answer from me. I was absent for two weeks and I have classes with students. In reference to your FAX of 5 April." "Thank you for advices connecting with sending SML script to your website. Also thank you for material (Getting Started booklet, some pages from Press Relise 5.9 [???]) but I don’t need them (because I have from your website–some days after you publish them on Internet." [New materials are downloaded by many from our web site as fast as we can create and publish them.] "Three years ago I was starting use TNTlite during my classes with students (my subject–Remote sensing and the basics of image processing). From the 5.8 version [and subsequent versions] I make a CD for students with TNTlite (for Windows environment) based on material from your website. This CD contains TNTlite system and data sets, all Getting Started booklets, Press Realise, SML and geoformula scripts, published papers, presentation’s materials (.mpg added for 5.9). Each student can take it, next install on his/her computer and learn more about GIS/IP." [I think that he means the Release MEMO, as our Press Releases would be of little value to his students.] "...Now about localization TNTmips into Polish. We planned this some time ago. Three person are connected with this task. But now we have no time, because we have lectures and classes with students. For the moment we prepared dictionary (version 6.0) for translating. We plan to finish this work before end of June (perhaps for the version 6.2)." From MicroImages dealers Extracted from email on 14 December 1998 [Using an early beta V6.10] "Localization. One of the first things I tried was the new localization tool. It works great and saves lots of time in comparison to the ‘work-around’ I had been using. Maybe you remember, for updating of the resource files I compared the versions with Textpad, import the result and both versions of the resource file to dbase, let dbase change the resource file and export to plain text format and then start translating new lines. Really very time consuming. The translation of the resource files is now underway. I calculate that it would take around 40 hours to update from 5.7 to 6.0." Extracted from email on 14 December 1998 [Using an early beta V6.10] "Looking around in some processes showed me that the staff of MicroImages again has done a great job. Now it is up to the community of all resellers, to convince people of the advantages of MicroImages products." Extracted from email on 11 January 1999 "I was pleased to hear that all the Ichthyology students received first class marks for their GIS-based projects. This is a credit to their hard work, our training and your software–remember these are non-geography students, who had never used any GIS system before. That they were able to get to grips with concepts of spatial data handling, analysis and presentation, while remaining focused on their project goals demonstrates how well-designed your products are, and how they are geared to productivity." Extracted from email on 11 January 1999 "Last week we received Version 6.0 of TNTmips. Congratulations! Although being delayed quite some bit, I consider this to be the most impressive quarterly upgrade package we ever received. Upgrade installation was smooth and the performance (both speed and memory footprint) appears to be significantly improved over earlier versions. The artwork you provide with the package is IMHO of excellent quality and the range of themes now covered by the Getting Started Booklets is–as far as I can see it–without example in the industry; I can’t see where in the GIS software industry the quality of layout and content of these little booklets can be matched. The SML functions you added will help us a lot with our ongoing development of precision farming applications based on the TNTmips platform." Extracted from email on 2 February 1999 "Well, I said ‘99 was going to be my SML year. I needed to write something for a project I’m working on, so I jumped in the deep end and got going. Last time I really programmed was during my [degree program]. Logic is a bit rusty these days. I found it tough going but within a week I had programmed what I needed, complete with selection widgets, result report forms, and fancy bit like coloured ‘lights’ that come on next to the result to inform the user of the quality of the result. In the process I got to grips with reading and writing files, working with widgets, georeference translation parameters, classes and members. Now, more than ever, I realize the power of SML, and really am excited about what could be done here. The SML I’ve written, will, with a bit more added in, be a very marketable product. Only problem is, its addictive! I’ve forgotten the sense of fulfillment one gets writing some code and getting a product that does just what you want. I got so hooked that I lost plenty of hours of sleep. Bottom line is, I’ve crossed the SML hurdle, and its really not that hard. I feel better knowing what it’s all about and that I can do it–can now honestly give clients better support here. Just one thing–I hate designing widgets–getting things in the right place and selecting the right kind of widget etc. is a pain, and takes up most of the programming effort? Your wizard idea, as implemented to design contour label queries, could be great to have here, i.e. a wizard to help one design forms/place widgets." [The introduction of an SML for Windows in V6.10 and its subsequent perfection will eventually allow the many Windows layout, design, and wizard tools to be used in creating scripts.] "... maybe I’m weird, but when the TNTmips box arrives here it’s like Christmas–I rip open the box to get at the Memo and no one can get any sense out of me until I’ve read it cover-to-cover. It goes to bed with me. Obviously, some people just go for the CD, install and carry on–and complain first." Extracted from email on 2 February 1999 "Yes, the course is over now and two more users are on their way. The two systems are being used on-site [in a project area]. After showing them the nuts and bolts of TNTmips, we got into their data and used TNTmips to model and evaluate one of the [areas]. They didn’t expect that TNTmips could do this–a bonus for them and they were delighted. The results we got confirmed some of their crude hand calculators, which they weren’t to confident about going to management with. Now they have some really good quantifiable and reproducible results, to immediately show management. These guys are really excited about the possibilities, and how quickly things can be done; analyses which took weeks or months to perform, should take less than a week now." Extracted from a FAX on 23 March 1999 "Some information about TNT products in [a small nation]. A lot of people requested me for information about TNTlite. I think that in [this small nation] are a lot of downloaded installations of TNTlite. My estimation more than 300 users." Extracted from an email on 15 April 1999 "Some days ago I downloaded the latest version of TNTdisp and TNTedit. The new legend view works great, its helpful for everyday GIS-jobs." From a prospective dealer Extracted from email on 22 February 1999 "We are authorized ESRI business partners, entitled to re-sell ArcView. Does this preclude us from MicroImages reseller program? (personally, I believe that MapInfo and ArcView are both analogous to high school level ability, whilst TNTmips is high school, university, through to PhD. level–basically it’s most usual limitation is with the users imagination or knowledge of the program.)" [This party was subsequently signed up as a new MicroImages Authorized Dealer.] From TNTlite users Posted on comp.infosystems.gis by Eric Miller on 23 November 1998 "After browsing this newsgroup for awhile, I wonder if anyone uses TNT-Mips/TNTlite? I’ve been using TNTlite for about a month and a half on my Linux machine, just to check it out. It has some really cool features that just smoke ArcView and are much easier than Arc/Info." "Haven’t used all of the features/aspects...I’ve been doing a little change detection studies with NALC data, and was pleased to find several variant formulas built in (though algebraically they’re pretty simple.). I couldn’t figure out how to do something similar with ArcView without writing Avenue code (belch!), Arc/Info’s easy enough in Grid but you still have to look up all those commands and parameters and then write AMLs...." "Anyway, I really dig the fact they (MicroImages) have a system that can run on Multiple Platforms *easily* and the files can be shared between Macs, Windoze, Unix’s without conversion. Also they don’t spread so many damn files all over the place, instead using a single RasterVectorCad file system. Their GUI needs an update, but hell it’s the ease of use and the output that matters. Anyway, I’m just looking for other peoples opinions on their product (not salespeople!)." Extracted from email from anonymous on 8 December 1998 "I just downloaded the Mac 68k version of TNTlite from your website." [continues on with question] "Just a quick aside, Congratulations on your documentation. I downloaded some of your booklets and was very impressed, both by the content and the style. I am also very impressed by the range of features included in you product. One last point, I think having a uniform product for all the platforms is a great idea!" Response posted on comp.infosystems.gis by Robert Sanson. govt.nz on 24 November 1998 "I have been trying out TNTlite for about the last month or so. It is a very comprehensive system (image processing, vector topological, CAD (non-topological), surface modeling), and it seems like they usually give you several algorithms to pick from for each process you want to run." "I’m still trying to get to grips with their database system though. Instead of providing functions for things like computing centroids of polygon features (like ArcView), they include them in an internal table (like Arc/Info), however I haven’t figured out how to get them from here into other tables (I’m not sure how they attach attributes to spatial features)." Response posted on comp.infosystems.gis by a MicroImages client on 29 November 1998 "TNTmips product are very, *very* good." "Another unsolicited testimonial. Just a mostly satisfied customer." Posted on comp.infosystems.gis by Orin Durey on 13 December 1998 "Anybody familiar with any GIS that will run on Linux? I saw a post that suggested GRASS. Any others?" Response from Russia posted on comp.infosystems.gis on 23 January 1999 "Check TNTlite (www.microimages.com) if you can download 100 Mb, you can get free trial version (although with some limitations)." Extracted from a posting on comp.infosystems.gis on 23 January 1999 "TNTmips will run on Macs using MI/X server. Check out http://www.micro-images.com. They have some impressive figures about the speed of their software on the new iMac. You can try our a free version, just needs a key to become fully operational, that’ll give you an idea of what you can do with this software. It’s a pretty powerful packages with good support for a variety of file formats. It does GIS, Image Analysis, CAD. The software isn’t exactly cheap though, so it may be more than you need."
An exchange on comp.infosystems.gis on 26 January 1999 [Question] "I was hoping one of you could help me with the following problem: I have two scanned images which in part cover the same area. I want to register one to another in such a way that I can use these images in multispectral analysis. In other words, I do not just want to align features (georeferencing) but I want to resample to the same grid so pixels match geometrically. I am using MicroImages’ TNTlite but I would also like to know if other software packages offer solutions for this kind of problem." [Response from a MicroImages client] "If you know the resolution of your input images (which you should), I think this can be done with TNTmips under the Extract menu. You can define areas from each image that ‘match’ and extract them. If you use the same sized area for both, and both images were scanned at the same resolution, then they should overlay properly–provided relief displacement and such aren’t issues. It may take some playing to get images to match properly, without georectifying them." Extracted from a posting from Canada on comp.infosystems.gis on 10 February 1999 "In addition there is TNTlite from MicroImages, free for download, but with some restrictions on the size of files, and import/export capabilities, but of a student it should not be a problem." Email from Canada on 19 February 1999 [Responding to MicroImages’ support] "Thanks Melanie. I’m using TNTlite 6.0 so that explains it. Great program!" Email from Canada on 10 March 1999 "As discussed earlier with you [via phone], I received the TNTlite software image processing and GIS and I think it is a very helpful tool for students and academic works in remote sensing and GIS. Since there is no French version of this software, and taking into account that remote sensing studies are getting more and more popular in high schools and universities, I was wondering about the possibilities of an agreement between MicroImages and [an academic institute] for the French translation of the software and tutorial. The [Institute] develops image processing algorithms (filtering, texture, segmentation, classification) and is also one of the most important center for remote sensing in Canada and among French speaking countries." Extracted from a posting from Texas on comp.infosystems.gis on 19 March 1999 [Question from Canada] "As a follow up to my last message, my focus is on which companies are targeting universities, colleges and technical institutes with their software. I am aware that ESRI seems to own the market in the States but am wondering who the underdogs may be and also has a presence in Canada and Mexico." "I am trying to compare various academic programs that each company offers, specially for the following software: Remote Sensing, Digital Photogrammetry, GIS, Mapping, Image Visualization. Thanks for any input, thanks for those that have already replied." [Response from Texas] "While they don’t target academic programs, MicroImages’ TNTmips (http://www/microimages.com) pretty much covers the topics you are interested in, and there is a student version called TNTlite (http://www/microimages.com/products/tntmipsfree/) which is very inexpensive ($10 for CD or $50 for CD and 1000 pages of tutorials). These programs do not have market dominance but do have an international distribution. In addition there is a student workbook that can be used with either program." Email from unknown on 19 April 1999 "I have encountered two problems/issues running TNTlite v6.0 on my Power Mac 7100/66 (32MB RAM)." [Two problems followed relative to using library spectral and endmembers in the hyperspectral analyses]. "Otherwise, the new hyperspectral analysis capability of TNT mips is quite powerful an well thought out. Bravo!" Posted from Australia on comp.infosystems.gis on 20 April 1999 "Hi all, Interested to know how many out there are into TNTmips, and what you think of it. I’m fairly new to it but am quite impressed with its range of capabilities. Especially impressed with TNTlite, price (free) documentation and datasets." "Is there such as thing as a discussion list or FAQ site?" [A TNTmips FAQ will be up by the time your read this.] Response from Florida posted on comp.infosystems.gis on 28 April 1999 "MicroImages’ website (www.microimages.com) has much info that is of a FAQ in nature. There was a short-lived discussion list, but it was terminated (don’t know why)." "As a user of both the old DOS-Based MIPS, and the current Win9x and Linux versions of this GIS package. I am extremely satisfied customer. The folks at MicroImages don’t spend a lot on advertising, but they do on continuous development of their software (FOUR upgrades/year)." Email from a computer science department in Japan on 22 April 1999 "I greatly appreciate permission to include your free product MI/X in our [student] CD-ROM. And, if you don’t mind, I would like to include TNTlite too. I didn’t know about TNTlite before, but now I have checked the web-page out and think it is really useful for our students. Well, I know you re also providing a free copy of CD-ROM including TNTlite. But it is difficult for us to distribute one more CD-ROM for all (over one hundred of) students [in computer science]. Anyway, it helps to make directory tree within our CD-ROM to send me a copy of your CD-ROM. Please send to the following address." 
©MicroImages, Inc. 2013 Published in the United States of America11th Floor - Sharp Tower, 206 South 13th Street, Lincoln NE 68508-2010 USA Business & Sales: (402)477-9554 Support: (402)477-9562 Fax: (402)477-9559 Business info@microimages.com Support support@microimages.com Web webmaster@microimages.com |