blankblank blank


Archive for the 'GPS' Category Page 4 of 24



Choosing A Laptop For A Cheap GeoPad

A GeoPad is a field-capable portable PC workstation configured for use with geographic applications (GPS/GIS); see yesterday’s post for more details. The GeoPad website has a full page on what to consider when selecting hardware, and is definitely worth a look. Much of that site is oriented around the purchase of standard field-ready laptops, none of which you could really call cheap. To put together a budget GeoPad, I concentrated on finding one that would meet as many of the following requirements as possible:

1. Relatively inexpensive; less than  $1000, preferably much less.

2. Durable. For standard laptops, 1-year laptop failure rates due to manufacturing defects run at about 10-15%, and 3-year at 15-25% under normal use (up to 31% overall after you factor in “user clumsiness”).  It’s likely that these failure rates would be greater for a laptop that gets jostled in a backpack, or used on a bumpy road. A mil-spec “rugged” PC would be the optimal choice, as it meets US military requirements for shock resistance, temperature, dust, etc.., but these are way too expensive.  A Panasonic Toughbook would be a great choice, but it’s $3000+ for even the most basic configuration. And you have to be careful about the word “rugged”, as there’s no standard definition. For example, the HP Elite 2730P calls itself rugged because it passes the military tests for temperature, dust and altitude, but not the impact/drop tests that reflect the kind of abuse it could see in the field.

3. Touch screen convertible PC. The GeoPad website emphasizes the utility of a touch screen interface for use in the field, and I totally agree. I’ve tried using my Acer Aspire One netbook in the field, with a clamshell configuration, and it just doesn’t cut it. To use it, you have to open it up, and enter data with the keyboard and mouse, something you can’t do standing up unless you have 3+ hands. A touch screen allows two-handed use at a minimum; combine it with a good working tablet case, and you can easily use it without dropping it.

Convertible means that you can use it either in standard clamshell mode, or rotate the screen and fold it down to use it in tablet mode. I wanted the option to use it in standard mode as well as tablet for added flexibility in data entry. You can get “slates”, tablets with no built-in keyboard, and then attach a keyboard via the USB connector; IMO, too clumsy and inconvenient in the field.

4. Good screen visibility in outdoor conditions. Most laptop screens don’t do well in outdoor conditions, especially in direct sunlight; the contrast gets washed out. Some rugged PCs come with special transmissive or transflective screens that will work well in outdoor conditions, but they can be very expensive. To reach my price range, I expected to have to compromise somewhat on this, but still wanted to find one that could be used under some outdoor lighting conditions.

5. Windows OS, because of the larger selection of free software available under this platform.

6. Respectable battery life; at least 4 hours under normal use.

7. Decent memory; 1 GB RAM, plus at least 30 GB storage space for software and data.

8. Light weight and compact size.

9. Plus as many typical notebook features as possible (USB ports, VGA output, wireless, Bluetooth, webcam, memory card slot, etc.)

The cost restriction narrowed down the field immediately; I couldn’t find a single “rugged” laptop available below $1000. Some standard touch screen laptops are available in that price range, in particular some of  the HP tx2 models which are just under $1000. But HP has the worst record for 1-year and 3-year failure rates in the industry (15% and 25%), and reviews have complained about the mediocre battery life. I was hopeful that some netbook touch screen convertibles might do the job, like the Asus T91 or Gigabyte 1028 series. While not “rugged” laptops, they offer a touch screen, 5-hour battery life, low weight, and fairly low cost ($500-700). But while they’re “touch screen” PCs, they have a drawback that would make them hard to work with in the field. A touch anywhere on the screen of these netbooks, even if it’s an accidental brush by your finger or palm, is registered as an input; if you’re trying to enter data on-screen, this can result in a lot of mistakes and frustration. More advanced tablet PCs get around this problem in one of two ways:

- A “digitizer” mode, where the only input recognized is that from a special stylus pen.

- “Palm rejection”, where the touch screen driver can recognize and reject minor contact from your palm or finger touches, but recognize firm contact.

So these cheap netbooks might not be a good choice because of this touch screen issue. The Hp tx2 models offer the option of being used in either basic touch screen or digitizer mode, but are much more expensive and might not be durable enough; at $1000, it seemed risky to go for this model. I almost gave up on the idea of a cheap GeoPad until, by accident, I found a fairly inexpensive laptop that seemed to meet most of my requirements. But that’s the next post in this series  …




A Cheap “GeoPad” – Putting Together An Affordable Field-Capable GIS/GPS Workstation

Last week, I posted my feature “wish list” for a handheld GPS unit designed for advanced fieldwork by scientists, mappers, technicians, etc.. One step up from that “dream” GPS unit are the mini PDA-like GPS devices / handheld computers already on the market from companies like Trimble, Ashtech, Leica Geosystems and many others. These typically have somewhat larger display screens, usually touch-enabled, and often run Windows Mobile. These can be a good choice for field data acquisition, but also have some drawbacks:

  • The screen size is fine for data acquisition, but can be too small for actual mapping work (both viewing and creating maps)
  • Even the cheapest units will start at over $600; more expensive units can easily run into thousands of dollars
  • The software for data acquisition and mapping is limited by the OS; there’s a lot less available for Windows Mobile than for standard versions of Windows (XP/Vista/7)
  • Windows Mobile isn’t the greatest OS in the world, and development of it has been very slow
  • The software that is available is often expensive
  • Processor speeds tend to be slow
  • Input is often by stylus only, which can be slow
  • Storage space for datasets can be severely limited

One step up (or maybe sideways) from these PDA devices is the concept of a GeoPad, a field-capable full PC running a desktop operating system like Windows XP. I’ve been keeping my eye on these for a while now, as this isn’t a new concept. The University of Michigan’s GeoPad website talks about developing the concept since 2003; they define a GeoPad as:

a rugged Tablet PC equipped with wireless networking, a portable GPS receiver, digital camera, microphone-headset, voice-recognition software, GIS software, and supporting, digital, geo-referenced data-sets.

The advantages of a GeoPad over a PDA-based solution (which the GeoPad site calls a GeoPocket) include:

  • Larger screen area, better suited for mapping, and easier for multiple people to view
  • Full Windows OS, which opens access to all Windows-compatible applications
  • Greater storage space (hard drive or SSD), for more datasets
  • More input options: full keyboard, stylus, mouse/trackpad

But it has disadvantages compared to PDA-based solutions as well:

  • Larger and heavier
  • Shorter battery life
  • Less rugged

But the biggest obstacle to GeoPad adoption might be cost. The GeoPad website lists a number of hardware/software combinations, with costs of about $4000-$5000; while the models are a bit out of data (circa 2007), costs of hardware and software comparable to the ones they list indicate that prices haven’t dropped as much as they have for other computer hardware.

About six months ago, I was asked to figure out whether it might be possible to find a combination of cheap hardware, and free/inexpensive GIS/GPS software, that would let you put together a GeoPad-type system for significantly less than the $4000-$5000 range without sacrificing too much in features. Over the next few months, I’ll be writing a series of posts on how I put together a usable GeoPad system for under $700 in hardware and software costs. It’s not a perfect system;  I’ll point out where it’s deficient, and how to work around some of those deficiencies. But you’ll find some of those deficiencies on expensive systems as well. Overall, my cheap GeoPad does most of what the more expensive systems do, at less than a fifth the cost. And even if you don’t go the cheapest route, hopefully you’ll find some of my experiences useful in putting together a GeoPad system of your own.

PS If you have no interest in this topic, don’t worry; I won’t be focusing exclusively on this. Posts on standard blog topics will continue.




In Search Of The Perfect Fieldwork GPS

So I’ve been following GPS Tracklog’s coverage of the Consumer Electronics Show last week in Las Vegas, and been underwhelmed by the handheld GPS announcements:

  • Magellan announced nothing new.
  • deLorme announced their PN-60 model; while it adds an interface with a Spot Satellite Communicator, the GPS capabilities are essentially the same as with the earlier PN-20/30/40 models.
  • And Garmin announced the Oregon 450 and 450T, which join the Oregon 200, 300, 400, 400T, 400i, 400c, 550, 550T. They all pretty much offer the same GPS functionality with minor option tweaks (electronic compass, barometer, camera, built-in map sets).

At the same time, I keep getting asked by field professionals what the best handheld GPS is for serious field work. I have to tell them that there isn’t a single model currently available that does everything I’d like to see in such a unit, so they have to make some compromises. While professional-grade units are available from Trimble and Ashtech (formerly Magellan Pro), they tend to be much more expensive than consumer units, less rugged, have shorter battery lives, and run Windows Mobile (aack). And while they’re good at data recording, they’re nowhere near as good as standard consumer GPS models when it comes to navigation and map display.

There must be tens of thousands of field workers out there who would be a ready market for a reasonably-priced GPS that met their needs, like geologists, biologists, archaeologists, etc.. I’ve put together a list of what I’d like to see in such a unit; feel free to add your own suggestions in the comments. And it’s not like the requirements are all that difficult to meet; in one form or another, you can find most of the following features already present in at least one commercially-available consumer-grade model.

  1. High-sensitivity, high-speed, high-accuracy GPS chipset.
  2. Galileo/GLONASS support for higher accuracy.
  3. Readout of the Dilution Of Precision, plus a graph that shows you the DOP for the rest of the day so that you can choose times of lowest DOP to get the highest accuracy with your GPS measurements.
  4. A high-gain, omni-directional antenna. Many units these days come with just a small ceramic patch antenna, and tend to have less signal gain and be more orientation-sensitive than the “stubby” quad-helix antennas found on models like the Garmin 60CSx and the Garmin Colorado series.
  5. A decent-sized screen that’s easily visible in direct sunlight. 2.6” diagonal is a bare minimum, 3” is even better. Many recent models have higher resolutions and greater color depth than older models, but sacrifice daylight visibility as a result. I’d rather have a unit with fewer colors and lower resolution that you could use in daylight than a sexy high-tech screen where you can’t see anything on it unless you hold it in the shade with the backlight on.
  6. An easy interface for adding text. If it doesn’t detract from screen visibility, a touch screen is fine, but a slide-out keyboard would work as well.
  7. Buttons with programmable functions. One of the big problems with touch screens is that you sometimes have to dive multiple menu levels into the interface to access a needed function; the ability to call up such a function with a single button push would make life a lot easier.
  8. Decent battery life, at least 15-20 hours.
  9. Rugged and waterproof.
  10. More waypoints, tracks and trackpoints. Some models currently sold allow only 1000 waypoints, and 10,000 trackpoints, to be saved. I’ve never had to have more than several hundred waypoints myself, but I’ve hit my unit’s limit of 20 tracks and 10,000 waypoints many times.
  11. The ability to add your own GIS map data, like point/line/area shapefiles, for display on the unit, preferably with at least some attribute data.
  12. Loadable vector maps, from the manufacturer and/or custom maps created by the user or others.
  13. Raster imagery like topographic maps or aerial imagery, both standard mapsets (like USGS topos) and your own custom imagery.
  14. A tri-axial electronic compass that works regardless of how you hold the GPS.
  15. If you have a tri-axial compass, you have the electronics necessary to determine the angle of orientation of the GPS with respect to the ground. This would be useful for geologists (dip and strike), archaeologists (site maps), geomorphologists (ground slopes), botanists (calculations of tree height using angle), and presumably others as well.
  16. Some means for recording additional data associated with a waypoint. Standard waypoints are limited to about 30 characters in the note field, and expanding that to a larger size to add more info would be helpful; another option would be audio recording capability that’s linked to the waypoing.
  17. A built-in camera with both automatic geotagging, and automatic tagging of the direction the picture was taken in. 5-megapixel minimum, with autofocus and a macro mode for closeups.
  18. Bluetooth NMEA serial output so you can use it with a laptop or PDA.

It’s unlikely that any of the current manufacturers of consumer-grade GPS models will someday decide to create a “prosumer” model with a reasonable subset of the features above. But a guy can dream ….

… and he can also try to put together an affordable field-ready PC with GPS and GIS functionality. More on that soon.




Free Routable Street Maps For Compatible Garmin GPS Units

For many Garmin GPS units, including most handheld models, the only mapsets that currently supports turn-by-turn routing are the City Navigator products, like the $75 City Navigator North America. These are available in two formats, both of which tie the maps to a single GPS unit. In the microSD card format, you install the card into the GPS unit, which means you can only use it in one unit at a time, and you can’t use it in conjunction with other mapsets; in fact, if you upload more maps to the card, you’ll erase the City Navigator maps! In DVD format, you can upload the City Navigator maps along with those from other mapsets, but the DVD maps are tied to a single GPS unit based on its serial number and a registration code – you can’t use them on any other unit, nor can you transfer them to a new unit you might buy to replace the old one.

mkgmap is a program that converts Open Street Map (OSM) data into Garmin-compatible maps, and it has recently added support for turn-by-turn routing. The Free routable maps website offers a convenient way to obtain these routable maps in several different formats.

Continue reading ‘Free Routable Street Maps For Compatible Garmin GPS Units’




G-Raster Updated: Raster Imagery In Garmin Handheld GPS Units

I’ve just released version 1.6 of G-Raster (alternate link if that one doesn’t work), my utility for creating raster KMZ overlays for use in some Garmin handheld GPS units (the Colorado, Oregon and Dakota series). Previous versions only worked with georeferenced imagery, like GeoTiffs,BSB, MRSID and raster imagery with worldfiles. For scanned maps without georeferencing, it’s possible to use free georeferencing software, but some might find that too technical. Starting with the latest version of G-Raster, you can use a scanned map that’s been georeferenced using Google Earth, a less complicated procedure:

  • Load the map image into Google Earth (Add => Image Overlay)
  • Use the map controls to move, stretch and rotate the map image until it lines up with the underlying imagery
  • Save it as a KMZ file
  • Load this KMZ file into G-Raster, which tiles the image and creates a Garmin-compatible KMZ file




New Utilities For mkgmap – Converting Open Street Map Data To Garmin Vector Format

Last year, I posted about using the free command-line utility mkgmap to convert Open Street Map (.osm) data into the .img format used by Garmin GPS units, with a later post trying it out with US OSM data based on TIGER data. If you need to access any of the command-line options, you’ll need to use it in its original command-line format, but in many cases you’ll only need to tell mkgmap which osm file you want to convert, a fairly simple command. I created a simple GUI to simplify this for the command-line-allergic like me, but there’s now a better one available from Activity Workshop. It’s open-source (written in Java), checks to make sure you’re not overwriting another file (which mine doesn’t), and also lets you combine multiple .img files into a single file:

mkgmapgui

Hopefully some of the command-line options will be added to a future version.

Also available from Activity Workshop is a program called osmwrangler. OSM files often contain not just road data, but also locations for amenities like restaurants, parks, doctors, theaters, prisons, etc.; a full list is available here. These will be converted by mkgmap into points on the Garmin map, regardless of whether you want them or not. osmwrangler gives you the option of removing whichever amenity types you don’t want in your map, and generating a new OSM file without them that you can then run through mkgmap. Oddly enough, it’s only currently available in command-line format.




G-Raster Version 1.3 Released – Raster Maps On Garmin GPS Units

I”ve posted about my free app G-Raster before – it converts raster imagery like topo maps and aerial imagery into the format compatible with recent Garmin GPS models (the Colorado, Oregon, and Dakota series). I’ve just uploaded version 1.3, which adds some major improvements:

- The option to rescale images that are too large for Garmin units (> 100 million pixels) to a smaller size; the previous version only let you crop the image

- Automatic rescaling of images to make sure that less than 100 tiles (the maximum for Garmin units) are created

- Display of the image resolution in meters per pixel, plus the maximum Garmin zoom level at which an image can be viewed without pixelization (aliasing)

- A view of the central part of the raster imagery, updated when the image is rescaled to a smaller size, or when the JPG compression is modified

- Your choice of 6 different algorithms for use in rescaling imagery

- Adjustment of the JPG compression factor using a slider, with immediate updating of the image view so that you can see the effects.

- Major additions to the help file

graster13




Tools To Help With Creating Garmin KMZ Files With G-Raster

Last post, I covered G-Raster, a free program I wrote that converts georeferenced imagery (GeoTiffs, MRSID, BSB, IMG, and graphics with worldfiles) into KMZ overlays compatible with some of Garmin’s more recent handheld GPS models (the Colorado, Oregon, and Dakota series). While G-Raster does a good job by itself on most such imagery, like that found at the USGS Seamless Server, there are a few programs that may be more convenient for obtaining georeferenced raster graphics (topo maps and aerial imagery), and modifying the imagery if necessary.

USAPhotoMaps: I’ve covered this free program before; it downloads USGS topographic maps, black-and-white 1-meter aerial imagery, and urban 0.25-meter color imagery from the Terraserver site. But you can also export JPG copies of the imagery using the Big JPEG function (File => Create a BIG jpeg file), which also exports a worldfile in UTM NAD83 coordinates. G-Raster has built-in support for these Big JPEG files; you only need to select “USAPhotoMaps Big JPEG” as the file type, then select the exported image file (found in the program directory after export). You’ll need the UTM zone as well – get that by choosing UTM coordinates for display (View => Lat/Lon => UTM). The imagery is in JPEG format, and won’t be in quite as high quality as that from the USGS Seamless Server, but should be more than good enough for most uses. As a bonus, you can also create GPX waypoint files directly in USAPhotoMaps using the map imagery; this lets you create both the navigation points and map imagery at the same time.

MicroDEM: I’ve also covered MicroDEM many times on this site. A small subset of its many functions can be helpful in editing/modifying imagery for use in creating Garmin KMZ overlays:

- The maximum pixel size for creating a Garmin KMZ overlay is 100 million pixels; if your image is larger than that, G-Raster will modify it to meet that criteria. But you can get better control over the final results by modifying the image yourself. MicroDEM lets you crop an image, or rescale it to a lower resolution.

- MicroDEM also offers direct access to Terraserver imagery, and can export it in GeoTiff format. While a bit more cumbersome than USAPhotoMaps, the final results are of slightly higher quality, and also offers the capability to choose the pixel resolution for the Terraserver imagery. For example, USAPhotoMaps exports black-and-white aerial imagery in either 1-meter per pixel or 8-meter per pixel resolutions; MicroDEM offers the choice of 1,2,4,8,16,32 and 64 meter per pixel resolution. By trading off resolution, you can cover larger areas in a single KMZ overlay file.

- Finally, MicroDEM can open shapefiles, color them thematically, and export the map as a G-Raster-compatible GeoTiff. While there are tools that can convert shapefiles into Garmin vector maps that display much more quickly than raster maps, this is a far easier approach, and gives you better control over the final appearance of the imagery as viewed on your GPS.