Elimination of motion artifacts in the pulse

Hi all! I have the task of implementing the algorithm eliminate motion artifacts when registering pulse. The task is to turn the forecast of the next pulse values on the basis of previous values. Curve pulse was formed window is the upper limit and lower. If the person is calm and artifacts there is no movement, then the range or window is comparatively large. If artefacts appear motion, i.e. very large/small values of the pulse, the range or window is drastically reduced. I know the function "Nonlinear Curve Fit.vi" which builds similar ranges of values. But this function does not suit me because it is very difficult to implement on the microcontroller. Because after implementation on LabVIEW I would like to implement this algorithm to a microcontroller. How do you implement this algorithm? Atmogramme with artifacts attached.
Attachments:
CRG.txt ‏2 KB

I agree that Lev-Mar is probably not a good choice for implementation on a microcontroller. It can be very computationally intensive.
Question regarding your text file: What do you expect the results to be for that file? It looks like peaks at 77, 90?, 99, 116, 134, 157, 171?, 195?, 211?, 236, 248, 257, and 266 may be the pulse signals you want. Peaks which may be at slightly different locations due to artifacts I marked with "?" The peaks at 195? and 211? are probably modified by artifact. No peaks before 77 can be obvioulsy identified as candidate pulses.
It appears that your data is undersampled. Many of the artifacts have only one sample. This means that much of the information in those parts of the signal may be lost. When I tried to look at a frequency spectrum of the data to see if it could discriminate between the pulses and the artifacts, the results seemed more random than meaningful.
Lynn

Similar Messages

Motion Artifacts or 3D like Problem in CS4!

Hello!
I have just noticed this strange footage in my timeline as the camera pans or the object in focus moves. The footage have got some edges or kind of shadows which follows behind like 3D footage without glasses or ghosting when the camera pans. This is not good for the eye when you watch it, and it just doesn't look right. Overall there is no smooth fluid motion in this footage where there is movement in the scene.
Original footage was recorded with a Canon XH A1 HDV 50i as I am in PAL region. I captured this footage using the camera's downconvert feature to SD. After capturing I purchased a Matrox RT X2 which I then used for multicam multiple monitors display. When I edited this footage with my client I didn't see this at all, and since it was a multicam project we had to cut between the cameras first then a second pass for refining and adding titles and transitions. All the editing was done as Matrox SD sequence. However when I played back the DVD thats when I noticed this strange motion artifacts in the footage. This is earlier in the year before the release of CS5.
I revisited the project because the client now wants more copies of the DVD. Although they didn't complain with this DVD I just thought I should correct the problem. The DVD is just 1hr 30 mins, so its not DVD compression here. This happens in the first half an hour of the DVD and then after that everything else plays smoothly and fluidly.Now the problem is that the footage in the Matrox sequence plays with the motion artifacts and it has a red line bar without any single effects. The othe footage does not have a red bar. But if I play it in the source monitor there is no such problem. This is happening on 3 out of the 8 tapes recorded on that event, and the tapes were recorded by one camera.
I have tried creating an Adobe PAL project but I cannot actually see as I haven't connected my preview monitor from my PC GPU! Still this SD footage have got a red line bar when in Adobe sequence. This includes even the footage that plays well.
Is this the matrox sequence causing this and if so how can I get rid of this problem without having to start editing from scratch!
SYSTEM
HP XW6600 Vista Biz 64
Xeon Quad 2.83Ghz
22GB RAM
GTX285 GPU (2GB)
PP CS4
RAID 0 for Source Files
Matrox RT X2
LG 22" monitor
Samsung 24" monitor

I wonder if this is a Field Reversal issue, as Matrox flips things around from regular DVD. That would be the first place that I would check.
Does this footage work in a non-Matrox Project?
Good luck,
Hunt

"Interlaced-like" motion artifacts when exporting

I am using CS4.
I rendered using the YouTubeWidescreen HD preset.
My original footage is from a FlipHD camera (H.264) and a JVC Piscio GC-FM2.
My intention was just to make some cuts to the in and out points, but otherwise not alter the orginal footage.
Two questions:
1. I have the right sequence settings (Sony XDCAM-X 720p), so that the FlipHD Camera footage comes into the timeline with any redline (or yellow line.) So, Premiere doesn't have to re-render the foogage to display it. But when I export the footage, it seems to (slowly) re-render everything. Back in the days of DV SD, I could (quickly) export DV footage without Premiere re-rendering it at all. Is there any way to do this with HD footage?
2. I am getting some sort of motion artifacts with the exported file. Looks a lot like interlacing to me. But the cameras shoot in progressive mode as far as I know.
The first photo is a screenshot of the original footage.
The second photo is a screenshot of the export video file.
The person on the left is spinning around.
In the original footage, each frame is without any motion artifact even though the person is moving quickly.
How do I fix this?

It turns out the JVC camera is at 59.94fps, not at 30fps that I had my sequence set to.
So, when I changed the sequence settings and export settings to match that, it fixed it!
Also, turning off frame blending also fixed it.
Is there any way to get Premiere to just export the original video without re-rendering/re-compressing?
This is way the way it worked with SD DV. If you didn't add any titles or any other effects, it would just export the original file (fast).
It seems I am losing some quality no matter how I export it.

FCE Exporting Nightmares...Motion artifact

I've done nearly a dozen projects now in FCE 4, but still can never get the same export format to work twice. My latest project: Filmed as always on a HDD camcorder 1920x1080, ingested and edited without a problem. Project looks great within Final Cut. (Duration about 4 minutes...There are a few clips filmed with a 640x480 underwater camera.)
When exporting using Export > QuickTime Movie... I get a 2GB file, but it contains bad side-to-side ghosting (double and triple images) during rapid motion. (NOT the small horizontal lines that I see on Google images of non-interlaced video.) I tried exporting it several different ways (using Export > Using QuickTime Conversion), all of which show this ghosting with rapid motion.
I finally took the first file into MPEG Streamclip. The same thing happened until I tried it is a .mp4 file. Now the resolution is slightly less, but the ghosting artifact is gone.
WHAT AM I DOING WRONG? Nearly every project I complete has some sort of artifact that I eventually get rid of after trying ten or eleven exports. The slight drop in resolution isn't nearly as appalling as the motion artifact. But I'd really like to know how I can export a project without motion artifact.

There are no changes in the motion tab.
This clip is probably the best example because my family is bouncing in a tube while being pulled by a boat.
Viewing in FCE4.0.1:
http://gallery.me.com/brutewolf#100011/3%20in%20a%20tube%20FCE&bgcolor=black
After exporting:
http://gallery.me.com/brutewolf#100011/3%20in%20a%20tub%20after%20export&bgcolor =black
The clip project properties include vid rate 29.97, frame size 1440 x 1080, upper field dominance, and ignore alpha.
It's difficult for me to tell if this is precisely the same frame (because of rapid motion, a transition, and no frame indicator in QuickTime) but it appears that the converted file is zoomed in with a slightly different aspect ratio. Is that a separate issue? The aspect ratio isn't terribly noticeable, but I think the shake artifact is.
This converted file was the one that was then converted in MPEG streamclip to a mp4 file (which eliminated the shake) and posted to YouTube here:
http://www.youtube.com/watch?v=Iguy39hTFps

Compressor creates artifacts at the beginning of clip. Then they go away...

So I make calculus videos and I having the file sizes small is important to me but when I compress my videos they have large artifacts in the beginning which then go away after about a minute. Any idea how to prevent this? If I raise the bit rate more they will go away but I would like to keep it low and i don't know why the artifacts go away after a while and it looks fine. Here is an example clip that shows what I'm talking about. Doing a multi-pass also helps a little but not completely.
http://www.guillermoteaching.com/filechute/intsublectureshort-medium%20200.mov
Any ideas?

but i want the motion in the beginning
Is there a way to do an adaptive compression that compresses the beginning at a slightly higher rate or is this just something I have to live with...

Importing Video creates motion artifacts

After days of frustration with Final Cut Pro X (I'm a new user), I've run into a really frustrating problem I've been unable to solve on my own:
I start with an AVCHD video in m2ts format, which I then "repackage" to .mov or .mp4. Then, I import into Final Cut Pro, and since FCPX doesn't play well with this format, I use the "optimize video" option to convert to ProRes. This takes awhile, but mostly works.
The problem I'm having is that the imported video contains motion artifacts. That's really the only way I can describe them. Little squares of the video that appear in the wrong place all over the video when something on the screen is moving a lot. If there's nothing much moving, no artifacts, but then as soon as something moves, there's little squares that appear/disappear all over the video and make it look awful. I've tried converting the m2ts to .mov or mp4, then importing, but nothing I can do seems to improve it at all.
I should mention that the original m2ts plays fine, the converted mp4 or mov plays fine, but once it's imported, it's borked. I thought maybe it was just FCPX, so I exported, but the exported video still contains these artifacts, so it's clearly the import process that is breaking these videos.
Does anyone have any suggestions for me? I'm really at my wits end about this

Sorry, let me be more clear.
I'm capturing video from a tuner card from my XBOX 360 in 720p at 60fps. The resulting video is an m2ts file which cannot be imported into FCPX. The app I use to capture has a function to convert to mp4, but I believe it's just repackaging from m2ts to .mp4, because it only takes a second or two to perform. Also, I use another application which losslessly converts to .mov. This also only takes a few seconds which means I'm pretty sure it's just the container changing, not the video format itself.
If I import this video as-is, without converting it to prores with "optimize video", the clip doesn't render, and there are green frames all over it, and sometimes it causes FCPX to crash outright. After doing a lot of research, this is a well-known problem as this video format is apparently "frameless" or something (I honestly can't remember what it's called). Anyway, it makes editing impossible. I know if I was using a camera I could use the import from camera feature, but because it's a tv tuner, I can't do that.
I've tried every playback format I can think of. Full screen, high quality, better performance etc, it's all the same. Also as I mentioned, even when I export it, watch it in VLC or quicktime it has the same artifacts.
EDIT: I should also mention that I'm at work, but I can post an example when I get home if that's required. I was hoping someone knew what I was talking about, but I can get an example too.
Thanks
Jeff

Panasonic Af-100 and Sony D7 in same timeline, motion artifacting

I'm cutting a piece together that has footage shot with both the Panasonic AF100 and the Sony D7. I got the footage from the AF100 first so I set my timeline up for that camera. Any of the footage from the D7 that that has pans/camera motion seems to have motion artifacting when cut into this timeline. Any tips on how to address this? media converter, timeline settings? I've used lots of different footage from different sources in the past and not had this problem, but I've never done a project using both of these cameras at the same time. Thanks for any help.....

Indeed. Also could be some of the CMOS roll on the D7. Do those shoot interlace? Not sure they would. But maybe that is what you are seeing?
As Shane says, frame rates need to match too or FCP will shove one into the other and it might not look right. Would look great in the viewer, but not so much in the canvas.
CaptM

I am trying to use motion tracking to have text follow a section of the background.The text box follows the motion path, but the text just sits there. What am I doing wrong?

I am trying to use motion tracking to have text follow a section of the background.The text box follows the motion path, but the text just sits there. What am I doing wrong?

Motion basically looks at the the source footage and it's timing (in the timing section of the Inspector) when analyzing for tracking. Filters are basically ignored by the tracker. Retiming actions, like time re-mapping, or retiming behaviors that are applied before the tracker should be used by the tracker. Did you change the timing after doing the tracking? If so this would also require you to re-track the shot. Understanding that you found a workaround, describing the effects you applied would help to explain why you ran into problems.
Cheers!

In Delayed Pulse Generation vi,Problem With THE PULSE WIDTH??

In Delayed Pulse Generation vi, I want to input a very low number for the Pulse Width while using an external timebase source. But the minimum pulse width has to be 2. Does anyone know how can I solve this problem??

Hey 45,
Unfortunately, there is no way to generate a pulse width smaller than 2x your external timebase.
There is an option to create a pulse of arbitrary width of your external source if you can afford some software processing in between. What you can do is use 1 counter to measure how many source edges of your card's internal timebase (80 MHz for TIO only, 20MHz or 100kHz for TIO and STC) your external signal is. This uses pulse width measurement as the counter application. Once you know how many source edges it takes to represent your pulse, then you can use triggered pulse generation and use the internal timebase with the pulse specs set to create the exact pulse width you want (and delay) and you can use your external pulse as the trigger. Th
is works well if your pulse is always the same width and you can measure it before hand. As an example, let's say your pulse is 20 internal timebase pulses when measured. This means you can use the pulse specs to specify a pulse width of 0.75 your pulse width by using only 15 internal timebase edges for your pulse width.
I don't know if I was clear above or not but if you give me your exact application you are looking to achieve, I might be able to help you out. Hope that helps.
Ron

How can I use the pulse train from a 6602 to trigger an niFGEN and niSCOPE on each rising edge of the pulse train?

Hello,
Here is my application: I need to use a 6602 counter/timer to generate a pulse train of certain frequency and duty cycle. On each rising edge of this pulse train, I need to output an arbitrary waveform on Ch. 0 of an niFGEN (5422) AND acquire data from CH. 0 of an niSCOPE (5124). I also need to synchronize the niFGEN and the niSCOPE to the same clock used for the pulse train (6602/ctr0). This process needs to continue until the user stops the system.
I can generate the pulse train using the 6602 just fine using ctr0, but the pulse train shows up on OUT0 by default. When setting up the niFGEN and niSCOPE to trigger on rising/positive slope edge, OUT0 is not an option for either device as a source for the digital rising edge (pulse train). The main options for both are PFI0-3 and RTSI lines.
Questions:
1.) Is there a way that I can direct the pulse train to a location (such as an RTSI line) where BOTH the niFGEN and the niSCOPE can use it as a start trigger for each rising edge? I noticed in MAX that a route can be made between ctr0's internal output and a trigger line and others. If this is a solution, could you please explain how to accomplish this?
2.) Once I configure the niFGEN and niSCOPE to be triggered on a digital rising edge, how can I effectively have this happen for every rising edge from the pulse train? In other words, can I just initiate the FGEN outside of the while loop and it will generate a waveform for each rising edge it sees at the source until the while loop is exitted?
3.) Is setting a reference clock for the niFGEN and the niSCOPE the same thing as synchronizing both devices using the same clock that generated the pulse train? It is not clear to me the difference, and why it would necessarily be useful.
Images of my current front panel and block diagram are attached. If you would rather have the actual VI's just let me know. Any help and/or explanation on this is greatly appreciated. Thanks in advance.
Attachments:
Front_Panel_Control.jpg ‏278 KB
Block_Diagram_Control.jpg ‏263 KB

Hello Cgifford,
Welcome to National Instruments Forums.
To output your signal to the PFI lines,
you can use external connectios between OUT0 and PFI lines. You can also use
the backplane to do so by routing into the same RTSI line.
1)
On the SCOPE and FGEN, the name of the
terminals are actually “PXI Trigger Line x/RTSIx” but on the 6602 you might
need to route the signal using the property:
You can also use the DAQmx route signal which perform the same opperation.
2)
This will depend on the frequency of
your pulse train. If this is lower than about 10 ms, then you can probably
place this on a loop and start and stop the acquisition every time. If the
frequency is higher than this, you will have to use:
- Scripting on the FGEN side (read more)
- MultiRecord Fetch (more information in the scope help file
section “Acquisition Functions Reading versus Fetching”).
3)
The short answer is yes. The longer one
might depend on how tight you need the synchronization to be (us, ns, ps). For
very tight synchronization, you should look into here.
Message Edited by Yardov on 06-18-2007 03:14 PM
Gerardo O.
RF Systems Engineering
National Instruments
Attachments:
property.JPG ‏7 KB

Being brand new to FCX I want to purchase Motion5 so as I am able to add a motion template to the start of a video. Can someone please explain in simple terms how I go about this. Looked at tutorials

Being brand new to FCPX and Motion 5 I want to use a motions template I have bought for a wedding DVD I have shot, so as to place the introductory template at the start of my film. I have looked at tutorials and cant seem to see anything that tells me simply how to achieve this. Do I edit my film as normal in FCPX and then transfer to Motion to add the motion template or visa versa ?? I need a simple answer as I am far from an expert. Can anyone help please.

what do you mean by 'motion template'?
as you notice on launch of M5, it offers you different project types.
when you select 'title', the minute you save it, it's avail on next launch of FCPX in the title-browser.
same with effects.
if you like to use M5 for producing a short segment (eg, I created my 'openers' with it, some animation of our soccer-team's logo, the first 4sec in this video), you simply export the final project as a Quicktime-file and add it to some Event in FCPX for future use ...

This morning when I turned my macbook pro on, an alarm sound came from my computer and a thin, short line of light was flashing next to the pulsing light that did not come on. The pulser light has not come on at all today.

This morning an alarm sound accompanied by a flashing thin light happened when I turned on my macbook pro. Now the pulser light is not coming on at all and there is a steady thin light to the left of the pulser that stays on when I close my laptop. Does anyone know what may have happened?

>an alarm sound accompanied by a flashing thin light<
That seems like a memory problem.
Try resetting SMC and PRAM:
iMac SMC and PRAM reset

Strange digital artifacts at the end of each song importing via iTunes

I've tried a few variations on archive searches and wasn't able to find an answer to this: In ripping some of my CD's into iTunes (either Apple Lossless or WAV) I've been getting some very odd, brief digital artifacts at the end of each songs on a CD. This sounds like two or three seconds of an audio tape on fast forward. It only seems to happen on new releases (I cannot think of any older CD's it's occurred on). I just got a copy of Mary Gauthier's new album (great CD by the way) and it happened on that one, for example. I tried ripping it multiple times in Apple Lossless, and WAV (always in iTunes) and each and every time it created files where this strange artifact existed at the end of every song. I play the CD I ripped the files from and the artifact is not there. I have several CD's, most of them recent releases over the past five years, that do the exact same thing. Otherwise everything else functions perfectly an the bulk of my 700+ CD library has imported without a hiccup. Is this some kind of anti-piracy technology...or do I simply have some setting off in my itunes preferences (Error-Correction is on, WAV or Apple Lossless is always set to Automatic, have experimented with importing to startup disk and external disk-same results). The artifacts seem to always sound the same too...they never vary, except that they sound different from song-to-song, but the same if say, the first cut is played over again. They are always 2 seconds plus or minus, and always at the end of a cut (the final 2 seconds or so). They do transfer verbatim to my iPod, and also play via iTunes through my stereo.
Anyone have any clues what this might be?

As a follow up, I went to the Genius Bar for a bowl of Mensa Stew, but only got some tepid broth. Here's what transpired:
The girl I was speaking with suggested that the artifact was the software misinterpreting information between the cuts. It was only a theory. She had not heard of any other incidents to relay. First she suggested I rip as MP3's. I explained that wasn't an option for me. She listened to a few cuts on a CD I'd brought with, or rather the ends of the cuts and confirmed there was a problem with the rip. She went through my settings, and asked if I'd tried it with Error Correction turned off. I explained I hadn't and that I preferred to use it. She suggested attempting to rip one of the cuts with it turned off just to see how it did. We did that and no artifact was there. So her 'solution' was to either leave error-correction turned off always, or turn it off for those problem CD's that produce the artifact (I'd estimate about 5% do on my laptop). This did not explain why it was doing it, nor why my g5 tower, which is running a carbon-copied system with all the same iTunes settings, does not produce the artifact with Error-Correction turned on.
I'd be interested to hear of anyone else who's been having problems importing hi-rez files in iTunes. I know of at least one other instance where the artifact produced (on a MacBook) is dis-similar to mine in that it takes the form of brief electronic beeps moving from channel to channel.
Anyone else?

HT4591 Why won't Motion give me the update to 5.0.7?

I'm running Snow Leopard and Motion 5.0.4. Software Update doesn't list anything for Motion, but the release notes say it's up to 5.0.7! Why won't Motion give me the update to 5.0.7?

Thanks! The problem I had was that I simply needed to download a new version from the App Store. I had honestly completely forgotten about the App Store. I'm on 5.0.7 now.

DAQmx - Measuring only the pulses w/o pauses of a pulse train

Hi everyone,
I've got a problem which I hope you can help me with.
I'm using a NI 4461 PXI DAQ-card on a PXI realtime system and LabView 8.5.
I'd like to measure a pulse train @ 1900 Hz with very low duty cycle. In order to reduce the aquired data, I want to measure not the whole train but only the pulses leaving out the pauses in between. I set up the task to aquire a certain amount of pulses (100) at the maximum sampling rate (204,8 kHz) with a definite numbers of samples per pulse (50) and using an available trigger line to trigger the measurement of the individual pulses. I thought to set the Read out butter size to (number of pulses) * (samples/pulse) = 5000 to be able to first measure the high frequency pulses and read them out after the number of pulses is reached.
But unfortunately it's not working at all.
Could you tell me, how to configure a task to do that job, please?
Thanx in advance,
Ralf

Hi Ralf,
Thanks for posting your question here. Are you still working on that issue? When you have done this, do you use the DAQmx-VIs or do you use a DAQmx-Assistant to configure the acquisition? There is also a website helping you to begin this ni.com/gettingstarted or this tutorial.
Feel free to post any other questions here regarding this topic.
Regards
TomBaum

Eliminatio​n of motion artifacts in the pulse

Similar Messages

Maybe you are looking for

Elimination of motion artifacts in the pulse