[Elphel-support] Fwd: expected imu data fortmat

support-list support-list at support.elphel.com
Thu Oct 16 14:40:20 PDT 2014


Hamid,

I would still use the full IMU rate and use some low-pass filtering - either in the camera or in post processing. I did not think the bandwidth will be a problem - simple averaging filtering would be easy add to the processing chain. It can be a separate program "piped" between the driver and output to the CF (now a simple "cat" command is used) that will average the IMU data. Raw IMU data is rather noisy, so averaging will definitely help.

Andrey

---- On Thu, 16 Oct 2014 14:21:34 -0700 Hamid M<mokh0006 at umn.edu> wrote ---- 

  Andrey,
 
 I was trying to lower the IMU rate to simplify the data collection process and keep things simple.  
 
 Sorry, by delayed response was to give it another attempt.  I've been trying to setup the camera so that:
 
 1) Plug a CF card (or USB into the camera)
 2) Upon powering camera, it will begin logging:
     - Jpeg Images at 3 Hz
     - GPS at 1 Hz
     - IMU at 50 Hz
     all into the medium of step 1 and will filenames or time-tags that can be corresponded to one another.
 
 Perhaps the flexibility of the elphel makes for the complexity that has made the above goal so far unreachable.  I'm not sure if the difficulty is due to improper settings or factors I'm overlooking.  
 
 Andrey, can I send you the camera to take a look?  Perhaps a refreshed firmware and your experienced eyes can really help us out.
 
 -Hamid
 
 
 
 On 10/11/2014 03:44 PM, support-list wrote:
 
 7da3289055.7045157628596740087 at support.elphel.com" type="cite"> Hello Hamid, Did you make the IMU work and now just trying to optimize parameters, or it does not work and you try to make it work at at least lower rate? If it does not work - please send the camera to us, we'll look at it, if you try to change the IMU rate - I am not sure if everything will work as expected as we tested IMU only at the maximal (nominal) rate. Why do you need to reduce the rate? Andrey ---- On Fri, 10 Oct 2014 14:22:55 -0700 Hamid M<mokh0006 at umn.edu> wrote ---- > Oleg, > > Just a friendly reminder - how do you reduce the imu rate? What string/number/hex is the 'imu_period' parameter expecting to achieve this? > > Thanks in advance. > -Hamid > > On 10/06/2014 04:23 PM, Hamid M wrote: > > 54330848.6020208 at umn.edu" type="cite"> Oleg, > > Right now the imu rate is about 2500 samples per second. Is the configuration parameter 'imu_period' used to reduce the rate of the imu? According to documentation: > > imu_period - set the IMU logging period in SPI SCLK periods, maybe > set to auto (default value, same as 0xff000000) to log when the IMU data is > ready. Value 0 turns the IMU logging off > > The mapping from imu_period to logging rate is unclear to me. For example, if you want 1 Hz do you set imu_period to 5e6? > > > -Hamid > > > On 10/06/2014 01:48 PM, Oleg wrote: > > bJYjvbWWNcw at mail.gmail.com" type="cite"> Hello Hamid, > > The parsed logged data looks ok - both, GPS and IMU. > > > * Fatal error: Maximum execution time of 30 seconds exceeded in /mnt/flash/html/read_imu_log.php on line 101 The timeout can be set in the PHP configuration > > > Please, let me know if you need any further help. > > Best regards, > Oleg Dzhimiev > Electronics Engineer > phone: +1 801 783 5555 x124 > Elphel, Inc. > > > -- Hamid Mokhtarzadeh http://www.tc.umn.edu/~mokh0006 > -- Hamid Mokhtarzadeh http://www.tc.umn.edu/~mokh0006 _______________________________________________ > Support-list mailing list > Support-list at support.elphel.com > http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com > >   
 -- Hamid Mokhtarzadeh http://www.tc.umn.edu/~mokh0006 _______________________________________________ 
Support-list mailing list 
Support-list at support.elphel.com 
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com 





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://support.elphel.com/pipermail/support-list_support.elphel.com/attachments/20141016/535c7d17/attachment-0002.html>


More information about the Support-list mailing list