Buffer of image in VBAI

Similar Questions

• Restore the Image of the camera target in VBAI

  I use Vision Builder AI 2012 with Smart camera NI1772C.  I can create target camera Image (Menu > target > create a target Image) but is it possible to restore the image on the camera?  The image is similar to PC save the image that can be safely stored and restored in case of emergency?

  Once you have created the image of the target, you can update any target of similar model with this image. VBAI launch and the start screen, select the target to update and do a right-click on the target and select "configure." On the last step of the configuration, you can navigate to the zip file that contains the image that you did to restore the target with all the image files.

  Hope that helps,

  Brad

• How to save an imaq image starting from a continuous stream, like a snapshot

  Hello

  I spent hours trying to find a solution for this on the internet for what I think should be really simple:

  I have a camera that gives me a stream of images, which I show in a chart of the image.

  Now, when the user clicks a button 'save', I want that picture that was shown at the moment where the user was hit 'save' is recorded. However, the user is prompted for a file location and name, so some time passes between hitting 'save' and write the image to disk.

  Meanwhile, the new images will be at the position of the pointer in the image that I wanted to save.

  It is clear to me why this happens, but how about him?

  I tried to make a copy copy Imaq, but this will give me just a new pointer that is always related to the old.

  I tried to convert it to an array, the same effect.

  I also tried writing a queue that contains a single element that has a buffer of image. The queue was not in a loop (which of course would replace the picture there), but again, the queue also just seemed to save the position of the pointer, while creating a new blank image in advance and I copied the image that I want with Imaq copy in this image buffer.

  So, how to store the actual image (not pointer) in memory to save a few seconds later?

  Basically I want something similar to what does the 'snapshot' of the curve of the image function.

  Any ideas?

  Thank you very much!
  Philipp

  Hello

  Thank you very much.

  I consider your solution already, however, I thought it's kinda an overdose.

  For some reason any, I don't know, he now works as I wanted. See example below (really just, I always wanted a picture).

  The only difference from what I tried earlier, is that now I did not specify the data type and the size of the new image... do not know if this is the reason why it did not work.

  Uday thank you very much for your help!

  Philipp

  PS: How did you make the image you posted in your message visible? Mine is just a link.

• IMAQ Overlay - more uniform than the image sensor

  Hi, I have a certain problem that could be more aesthetic than problematic.

  IMAQdx Grab runs in a loop, acquisition of the images in real time, which are then treated in this loop, and finally some calculated the points are superimposed on the image. The last thing I've done in this loop is to send the picture to overlay to an image of front control, so that the user can see the results superimposed on the image in real time.

  However, the image is mainly image appears without results superimposed so they appear only once in a while, very "flickery,". However, the results are well calculated for each image. If I use a probe to look at the purple image wire just before moving the witness in the block diagram, the probe constantly displays the lines superimposed flicker.

  I hope that I was a semi-claire, or see the attached picture. Does anyone have any idea why this might be?

  Thanks in advance

  Where did you clear collection to remove the previous measures? It is best to place before new overlay to labview more time to show.

  Maybe I'm wrong, but indicator can be only one command to display image and design occurs in the separate thread later.

  And probe creates the copy of the image data when it is called and did not progress until she's copying finishes. I prefer to do it manually with image: have a separate image display buffer (Create image.vi) and copy after all treatments and overlay is made (copy image.vi).

• VBAI IMAQdx to acquire initialization sometimes falls down and the ruins of my throat

  I have a GigE (FLIR A325) camera from here acquire images in VBAI.  I have several saved inspections and I use the selection of Inspection step to automatically switch between them based on a shared Variable.  I notice that there is an initialization that is performed after each pass an inspection on foot to the other.  Opportunity (actually about 25% of the time) the initialization fails.  The error is always related to the attributes of the camera (could not read or check... something like that).  The thing is that if I force the initialization of repeat by manually opening an inspection, it usually works.  The problem is that as soon as there is an initialization failed, VBAI gives up and stops at the inspection.

  Is there a way to force VBAI repeat the initialization on a mistake?  Of course, it would be great if I have no error at all, but I'm afraid that this will be a game between us, NOR and FLIR of finger.

  The main concern here is that when VBAI is running in automatic mode, it takes a method to recover errors like that.  When the inspection is stopped, the system just sits there idle and helplessly.

  Thank you

  Dan press

  PrimeTest Automation

  Hey Dan,

  Unfortunately, it seems that camera does not correctly handle definition of attributes (I guess I'll start the finger pointing). You change any attribute of camera during the inspection? You have several inspection with different parameters? If so, the camera could be in a different State then when you started, and when we close and try to apply the new settings to the step of the acquisition of the re-inspection, the camera does not allow for some of the attributes to be defined, because the current mode is not valid, and they have a problem with the order of things. A debugging tool that will identify the attributes of boredom is WindDbg

  http://www.Microsoft.com/whdc/devtools/debugging/installx86.mspx#a

  After this installation, you can add an entry to your regitry which turns on all the driver reads/writes about the attributes with status information:

  HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\niimaqdxk\Parameters

  Add "AttributeTraceFlags" as a DWORD value

  The attribute is a bitmask with the following values:

  //--------------------------------------------------------------------------

  Operations of attribute we can draw

  //--------------------------------------------------------------------------

  TraceOperationMask = 0x000F,

  TraceGet = 0 x 0001,

  TraceSet = 0 x 0002,

  TraceValidate = 0 x 0004,

  //--------------------------------------------------------------------------

  Status codes to track

  //--------------------------------------------------------------------------

  TraceSuccessMask = 0x00F0,

  TraceSuccess = 0 x 0010,

  TraceFailure = 0 x 0020,

  TraceWarning = 0 x 0040,

  //--------------------------------------------------------------------------

  Follow the exit to destinations

  //--------------------------------------------------------------------------

  TraceDestMask = 0x0F00,

  TraceDebugger = 0 x 0100,

  Trace file = 0 x 0200,

  Just so set to 0x01ff to see in the debugger. Once you find what is the problem attribute, you can speak with FLIR to see if they can fix their XML file, or try to make sure that the attribute is in good condition before closing... I would recommend talking with FLIR so everyone can benefit... FLIR can come into contact with OR and we can also help to ensure that their XML is valid if they need assistance.

  Hope this helps,

  Brad

• How to get the background color of a layer with AE SDK

  Hello

  I try to get the background color for my current layer, but I can't seem to find how do.

  I don't want background color, background color of the application, but the color that shows through, if I set the alpha of my pixels to 0.

  For example, assume that the background of the composition is black, I put a solid top, green then put a picture on top of the solid and apply my pug-in the image.

  If I set the alpha to zero in my plug-in, green solid will be displayed.

  So I want to be able to know that my background is green.

  I tried to use PF_AppGetBgColor(), but which returns the background color of the application, which isn't green in this example.

  If you want more details on what I want to do, I plug P1 which affects some 0-alpha pixels. Above it, I apply an another plugin P2 which reveals alpha-0 pixels by mixing with little color.

  What is happening now is that the output is color of the original image, mixed with the color of P2 (because only modified P1 alpha pixels).

  What I want is to have the color through the transparent pixels of P1 display mixed with the color of P2.

  To do this, I would like to be able to set the color of pixels 0-alpha in the color of the background in P1.

  Sorry for the question long.

  Thanks for your suggestions.

  Hello nicolas.

  If I understand it, that you try to get the buffer of image layers composited under the layer with your effect.

  If that is correct, then I fear that it is almost impossible to do.

  When AE makes a composition he doesn't it down. He's trying to do the opposite in fact.

  Why? because if at some point, the buffer is completely opaque, then he has no need to make what's below.

  further more, checked the diaper changing the rendering order, and many other things affect the rendering of the layers order.

  In addition, there is no API to access buffers intermediaries comp. just the end result.

  the closest you can get than without turning of the world, is to use AEGP_GetCompBGColor().

  that would be useless to you that the image is not composite on that color, unless it is the model being saved to the file.

  so why I said 'almost impossible '?

  You can transform the world.

  You can write a type ECAP 'craftsman '. (look at the example of 'arti')

  These plug-ins make comps (instead of the "advanced3D" converter)

  a craftsman has results intermediaries comp, because he is the one doing the rendering.

  so now you can have your craftsman save the buffer, you need and deliver it to your effect.

  I must warn you that this is very very very difficult.

  Another possible solution is to create a copy of the model resides in your effect,

  erase all the layers that has the image you need and make that reproduces model using AEGP_GetReceiptWorld().

  If you use this method, you must supervise the original model and apply changes to the copy.

  Yes, it's too difficult to do.

  you could make a change in the strategy and apply your effect to an adjustment layer instead.

  the entry for the adjustment layer is the buffer underlying layers.

  Then, you could get the original sources you want to treat using checked layer params.

  the last solution I can think uses the expression of sampleImage() on a hidden parameter.

  This will give you any layer's pixel data in the model, but not the composite of the underlying layers.

  Maybe if you tried sampling adjustment layers post effects... I don't know. you need to try.

  This method is very slow and is also limited to expire, so you can't get a large number of pixels.

  the expression simply fails.

  I hope that this was not all bad news for you.

  Maybe you said more about the plug-in you develop I could come up with a better plan.

  :-)

• On the use of sequence_data newbie question

  Hi all

  I am writing my first AE plugin effect, porting the existing stand-alone C++ code.

  Mine is a time filter that works on the current frame with the neighbors of +-N, and I don't mean to make a port full of code, but rather managed to operate as an effect with the fewest possible changes.

  For this, I copy data from frame of PF_EffectWorld objects to the type of image I'm using (I managed to do) and the Bank of consecutive in a circular buffer also images from my own fault. The memory for the buffer to be allocated only once, and raster data must be accessible from the rendering of an image to another; otherwise the rendering would be extremely little profitable.

  My question is if I can use the sequence_data field to store this circular buffer. I'm quite confused, read the docs, where sequence_data is related to sequence flattening of persistence, but being new to After Effects I have not understood what has been explained. Can I just use PF_NEW_HANDLE when I get a PF_Cmd_SEQUENCE_SETUP and affect my buffer uncontrolled in the field of the sequence_data? Is it safe to use this field I mean?

  Those who were my questions, but if you are already experienced with AE and think what I'm trying to do is fundamentally wrong because it goes against the programming/design guidelines/conventions/whatever of AE please give me any suggestions that you could save. I would be very pleased them.

  Thank you all for your time.

  so... sequence_data, huh?

  first of all, Yes. You can use sequence_data for what you described.

  You can allocate and dispose of this memory anytime you want.

  different appeals of the sequence data will simply tell you what happens.

  Allow me to explain. These are the calls you're going to get:

  PF_Cmd_SEQUENCE_SETUP.

  you get this call only once, when a new instance is first applied to a layer.

  you do not HAVE to allocate memory now, but it is the notification you get to have a new instance.

  If you do not allocate the memory now, you can later check whether or not a handle exists and do it whenever you see fit.

  How do you check?

  If in_data-> sequence_data is NULL, then no memory has been allocated.

  PF_Cmd_SEQUENCE_RESETUP.

  you get it after the following:

  When project loads with an existing instance of your effect,

  After copy and paste of your effect.

  After you save the project.

  now, you can decide if you keep the data as it is, or trash and start from scratch.

  PF_Cmd_SEQUENCE_FLATTEN.

  preparation for the Save/copy.

  you get this call only if you have requested is using PF_OutFlag_SEQUENCE_DATA_NEEDS_FLATTENING.

  I think it is of interest to you.

  Why? because if in_data-> sequence_data is something, AE records regardless of the handle for the project points.

  so if you have allocated 100MB to contian 50 frames in each direction, the project file will save these 100 MB within himself.

  on the other hand, if in_data-> sequence_data is NULL, then AE stores nothing.

  PF_Cmd_SEQUENCE_SETDOWN.

  all the garbage.

  this instance is being deleted. (or the project of closing)

  so, as you can see, you do not have to allocate / free at any time specific memory (except maybe hollow...).

  These calls are all logical times for some operations.

  You can allocate memory during installation of chassis or rendering, or when you want.

  the things I am concerned, using this method, are not technical. they are logical.

  first of all, if you're caching the images inside your effect, are you sure that these images are valid?

  Maybe the user has changed these frameworks? You can't really tell.

  Another concern is that you will not necessarily acquire these images in the correct order (if it is of any importance to you).

  you user might be jumping between two random images, then you will be accumulating non sequential images.

  It's may be solvable if keep you track of the timestamp origin these images.

  and finally, you don't know what random of washing is completed and made sequential began.

  There is an effect that makes this kind of cached image describe you: 'CC wide time FX '.

  It creates an echo by registering previously acquired images, instead of rendering each image.

  the effect is such a pain to use, it's just not worth it.

  When you scrub through the timeline the result is incorrect.

  and before rendering, it must look to each instance of the effect and click on 'options' to clear the buffer.

  not to mention several computers rendered is off limits, as each machine accumulates a different buffer, because jumps frames.

  Consider this:

  that plugin was created 14 years ago to when the fastest machine was about P1 166 mhz and 128 MB ram are all over the world.

  Nowadays, AE hiding executives, asking if the latest images every time doesn't mean necessarily that they made a new. they are very probably cached.

  just a thought.

• VBAI Image by 90 degrees for display rotation

  VBAI we want to quickly turn an image 90 degrees strictly for exhibition purposes.

  It is an image that is acquired from a GigE Vision camera.

  It seems the only option of rotation in VBAI is the Vision Gometry Wizard tool, which unfortunately does not update the horizontal and vertical image size, clipping out of the segment on the right of the image.

  Is there an another, better way to rotate an image 90 degrees such as the width becomes height and height becomes the width of the new image?

  Thanks in advance!

  Nelson

  Hello

  Here's a LLB containing "VBAI Rotate.vi", that you can call from the LabVIEW execution step.

  The Masker is sure of the value of the Src of the Image on the current image and size to maintain to False.

  Hope this helps,

  Christophe

• VBAI acquire multiple images

  Hello

  We work with VBAI 2012 and we want to acquire multiple images as quickly as possible one after another to decide what is best for the inspection. Is it possible to take for example 5 images one after the other, or what I call the image to acquire five times?

  Thank you

  Oliver

  Why not continuous use to acquire images and use 5 as the number of buffers required?

• Can I have a global calibration for different images sizes in VBAI

  I run a VBAI algorithm on images from several cameras to linear scan.  We use a square, simple, calibration (.42mm/pixel) for each image, they are always 1024 pixels wide, but can vary in length of the image.  Currently, I use several instances of Calibration Wizard then 10 to account for variations in length, but it seems to me that the internal calibration resulting is the same for each instance.  So is it possible that I can set up a unique calibration use if for any length of the image?

  Thank you

  This is a feature that we added in the latest version of Vision Builder AI (2012), which will be available very soon.

  Would it be possible for you to use the LabVIEW execution step to calibrate your image? If so, you can call a LabVIEW VI that sets the calibration using IMAQ Calibration2 Simple to set, after the steps in the acquisition. Once the image is calibrated, all subsequent steps must return results calibrated, even if you do not use the VBAI built in calibration.

  Christophe

• VBAI 2012 "save the image calibrated to the file" issue

  Hello

  Under the State of 'Calibrate Image', 'Save the image calibrated to the file' option does not work correctly.

  With the exception of the type 'Point of Distance' of the calibration, the new image is ever recorded. I checked the calibration image .cal2 in the folder \ProgramData\National Instruments\Vision AI\Calibration Builder, each iteration the meta-data are updated but not the image.

  I have a camera calibrated with a grid (see picture) and using the model of the distortion. Working remotely can sometimes change and I need to take a recalibration. This step should be as simple as possible. Example: put a model calibration in the field of view, then press a button recalibrate.

  The problem mentioned above, it is impossible. Whenever I have to manually go into the choice of the State of the Image calibrate and update the calibration image myself. Otherwise, calibration is recalculated from the original image which is not best suited for the new working distance.

  Is there an easy way to make VBAI save a new image of calibration with each iteration?

  Or is it something that I did not understand this option 'save image calibrated to the file?

  Vision Builder for Automated Inspection VBAI 2012

  Windows 7 64 bit

  Best,

  Ken

  Hi Larpin,

  Unfortunately VBAI does not support the suggested behavior. The best solution in my opinion would be to implement your application completely in LabVIEW!

  To Eval download link. Version:

  https://lumen.NI.com/nicif/en/evallv/content.XHTML

  Kind regards

  MISIC

• VBAI Interface - Get Image-> Dest Image inspection

  I have a complete Vision algorithm built in Vision Builder AI and programmed a VI in Labview that started the algorithm and results used to calculate. It works very well. Connection between Labview and VBAI is correct and the results are updated. Now, I want to show the image of the VBAI in my Labview GUI using the VI 'VBAI Interface - get Image Inspection' of the Vision Builder AI toolbox. Everything I try to connect at the entrance of "Image transfer" of the VI I am unable to let it run and give an image appears. See attached screenshot for details.

  -What type of entry how can I put a link to get this working VI?

  -Where can I find more detailed information in the context-sensitive help on all these VBAI VI toolkit?

  You must use IMAQ create to create an image. You can see an example of this in the case of Vision Builder AI under API Examples\LabVIEW Examples\LabVIEW API example (local) .vi

  Hope this helps,

  Brad

• Problem reading video image from the camera IP Axis - confusion of variant data buffer.

  Hi there;

  I am writing a VI for an Axis IP camera.  He went to work, but I need to extract individual video stream images so I can overlay text using the functions of text IMAQ overlay.

  The thing is when I take a picture using the axis dll library (GetCurrentFrame), it returns a buffer size and a Variant representing the raster data.  The data is = 921 640 bytes representing a 40 byte header, and then the 640 x 480 x 3 raster data.  When I run the vi, I get a correct value for the size of the buffer (921 640) so I know it works.

  What don't understand me, it's the video image is placed in a buffer of type variant.   I don't quite know how to extract the raster data of the data type of "buffer" which is a type variant.  My apologies, I'm a bit ignorant in how to handle the types of "Variant".

  Does anyone have a suggestion?

  This is the VI

  Hi Peter,.

  Just to confirm, what IMAQdx version do you use?  You want to get the version 2010.3 since it is the latest version: http://joule.ni.com/nidu/cds/view/p/id/1641/lang/en

  The next issue would be that the Ethernet cameras discover is not necessary and that it is only used for GigE Vision cameras. The VI was named before IP camera support has been added and the name is unfortunately confusing now. In any case, the discovery of cameras takes place in the background and is automatic. The VI list must list your camera. It may be worth trying just to see if the camera appears in first MAX. Note that the IP camera must be installed on your local subnet, so it can be discovered.

  Regarding the examples, virtually none of the IMAQdx examples included in help-> find examples should work.

  Eric

• How to convert FLIR gray Image to a map of temperature in NI VBAI

  I worked with an infrared camera FLIR A315 and software VBAI of NEITHER.  I need determine the presence and continuity of a hot mastic foam which has been injected into the hollow bulb of a bead attached to 4 adjacent places that flow together inside the bulb of weatherstripping to form a continuous rubber scoured the filler plug.

  So far, I was able to trigger image acquisition in VBAI from a FLIR infrared camera monitoring the status bit an Allen-Bradley PLC.   Communication with the camera FLIR A315 is via the IMAQdx GigE driver.  However, I did not understand how to determine the temperatures of the gray scale images I get from the FLIR camera.  I wonder if someone could give me some advice on how to do it.

  I have attached a few images below to illustrate what I work with and to what extent I'm away.

  • iPhone_Photo.jpg - shows a photograph of the stripping section that the FLIR infrared camera is 'looking '.
  • FLIR_IR_Monitor.jpg - a screenshot of a FLIR Infrared Monitor utility that shows the temperatures measured at a number of points and areas (showing that it is possible)
  • NI_VBAI.jpg - part of my VBAI 2010 project that shows the image, I have acquired the FLIR A315 IR camera

  So, I am part of the path, but so far have not found a way to map the image of gray-scale that VBAI acquires the infrared camera FLIR in a heat map where I can determine the temperatures - and would be grateful for the help on how to do it.

  Kind regards

  Nick

  Hello

  Image provided by the FLIR camera are unsigned 16-bit.

  According to the temperature scale FLIR selected (10mK or 100mK), you have to divide by a factor of 10 or 100 to get the values of K. Do not select Exit radiometric.

  If your object is cold, the population of the histogram of the image will be on the left side of it and if the object is hot, the population of the histogram will be on the right side.

  If you want that your values of temperature in degrees Celsius, just add 273.15 Kelvin value.

  Then you can add a color palette in VBAI (iron FLIR is available)

  Hope this helps

• How to write the raw buffer grabber (table 2D) image to PC memory?

  Hello

  I use a library C++ and NIImaq to save table raw image 2D on PC.

  I have a PCI-1428 acquisition card.

  Now, I'm writing a 2D framebuffer grab table memory PC I assigned using virtualalloc.

  Any way to do it?

  Thank you

  Travis

  Hi Travis,

  After further research, I found that you can use the imgSessionCopyBuffer function (see the reference help function OR-IMAQ for more information) to transfer your data from the memory embedded in your designated PC buffer.

  Also, you can find more information on how to set up in the developer zone article below Acquisition ring (Versus the system of onboard memory section memory speak your specific question).

  DevZone: Ring Acquisition

  Hope that helps!

  See you soon,.