Well I guess the day had to come, but sadly (particularly after the Gingerbread upgrade) my San Fran passed away…. well mostly. All of a sudden reception became very bad unless I was actually underneath a transmitter and it looks like the internal (GSM) antenna was damaged. O well…..
…..so two days later I took delivery of a shiny new San Francisco II. Basically the same unit (Google Play identify it as a Blade S) but with an uprated 800MHz CPU, 5MP camera (with flash) and forward facing camera (good for Skype). It ships with Gingerbread, but locked of course. One of the online sellers producers a good guide to unlocking your phone which I ran through and it worked (rooting subsequently required if you replace with a stock unlocked ROM). After which I then realised that the far simpler thing would have been to install a TPT for my phone which re-partitioned the memory and installed ClockWordMod to allow me to flash a new ROM. This was fast and simple. The main forum has a few ROMs available (there aren’t a vast selection) including Ice Cream Sandwich (v4.0) but I think Gingerbread is a bit faster so plumped for Fish ‘n Chips which works rather well.
The phone just works…..switching over was painless if a little time consuming in parts installing software on the phone. However being able to just switch the microSD card across is a boon. Lets hope this one last 2 years.
A nice cover image and article over at PE&RS celebrating 40 years of Landsat missions. The cover article looks at the analysis of the first ever Landsat scene which is a pleasant bit of nostalgia - and perhaps provides an excellent example of the power of multispectral remote sensing in identifying surface features.
Intriguingly Intermap has just release another global DEM in the form of the NEXTMap World 30. Perhaps the most useful document to read is the technical review which contains a fair bit of detail on the product - that is, its a fusion of ASTER GDEM, SRTM and GTOPO. They clearly believe that their experience in fusing data products has led to the production of the “best available” near-global DEM under the NEXTMap badge. I’m not going to argue and it’ll be interesting to see how many people start using this.
Well I finally took the plunge and, 18 months on from installed Froyo (Android 2.1), finally upgrade to Gingerbread (Android 2.3.5). Having spent a long time getting all the apps set up as I wanted and repartioning the memory to gain extra app space, I was loathe to replace what was a very stable setup. Anyway, the lure of an update got the better of me and my first port of call was the list of available ROMs (and the install guide for the San Francisco). This actually showed an extremely limited list of available ROMs for the first generation San Francisco, largely restricted to Cyanogen Mod. Clearly this list isn’t that up to date and visiting the ROM Forum is a better bet. I plumped for Swedish Snow which is a relatively lightweight 2.3.5 ROM - it is for second generation phones but, crucially, uses the Total Phone Transfer System (see install guide) which does the entire upgrade in one hit, BOOT, RECOVERY and ROM all at once, as well as the ability to change partitions. For Swedish Snow it upgrades a generation 1 phone memory layout to generation 2 and then installs the full OS. All in about a minute and then boot up. The only checks to do is that you have a generation 1 phone and that the downloaded ZIP file is not corrupt (MD5 checker).
And…. it just works! The ROM is well packaged, useful selection of apps and, so far, very stable.
Point clouds are becoming big news - as LiDAR systems move out of the niche and in to the mainstream, software processing shifts from high cost bespoke to lower cost and open source. And skills in these areas become desirable - at Kingston we run a High Definition Surveying module based in-part around our Leica terrestrial laser scanner where students learn about data acquisition and data modelling. Traditional areas have been those dealing with detailed infrastructure such as internal facilities modelling and management and highways/railways, but we have seen use in other areas have costs have dropped including crime scene/traffic accident and pylon monitoring. It always surprises me the increasing number of applications and this is well demonstrated through Radiohead’s House of Card video (including Google code, downloads and use of Processing.
Anyway, back to the point of the post which is the increasing proliferation of tools and there are a few points worth looking at:
-there is an international data exchange format ASTM E807 and an open source library -PDAL for the reading, writing, and translating point cloud data -PCL is for pointcloud processing including “filtering, feature estimation, surface reconstruction, registration, model fitting and segmentation, as well as higher level tools for performing mapping and object recognition.” -point-cloud-tools comes out of an academic research project and is focused on processing massive pointclouds (initially of fluvial environments) to calculate per-cell summary statistics-LASTools for point cloud converting, filtering, viewing, gridding, and compressing
How about 100,000?? A really great TED talk from Peter Norvig on how they turned their on-site classroom in to a 100,000 strong virtual class and flipped many of the ideas about distance learning to motivate and engage students.
OA costs raised its head again this week when I was asked to review a paper for Remote Sensing, published by MDPI who have specialised in open access for a number of years. I therefore wondered what their revenue was so looked at last years publications: 130 articles, spanning 2726 pages. MDPI charge variable amounts depending on journal (strange as I would expect costs to be the same!), with Remote Sensing at ~£500. That’s a tidy £65,000 from OA publishing. Of course their article numbers are healthy (2700 pages is good by anybody’s standards) which means (assuming all authors are paying) they can generate this income.
So no quibble that OA works, although it would be interesting to know what production costs are like. There is an additional cost for more significant language copy editing and their peer-review computer system is far simpler than available from T&F or Elsevier. I also suspect they don’t really have a marketing budget. If they were publishing 20-40 articles a year then the financial case would be more limited. So again, it is sector dependent. Can you generate the submissions and are authors willing to pay. Clearly £500-1000 is at the cheap of the market, with £1,500+ a more realistic cost for a mainstream publisher.
The Leaf Motion is doing the rounds, but is well worth repeating again anyway, not least because, amongst other things, it shows some Google Earth navigation. Minority Report style motion control is arriving….