Animate Particles Using Image Source Alpha Mask - Outline

I am trying to animate particles emitted from the outline of an alpha mask. I can get the emitter to see the mask, but only as a rectangular bounding box. How do I get the particles to be emitted from the organic shape within the mask? I created the alpha mask in Motion.

Heya,
I see what you're talking about: using the alpha for Fll or Random Fill seems to work fine, but using the alpha with Outline doesn't seem to work (except using the outline of the whole bounding box).
Maybe a bug, though I would guess that generating an outline path from an alpha channel would be an expensive process, computing-wise. It makes sense if you have a relatively simple alpha with a single "blob" of solidity, but if you had a complex (noisy) alpha, it would be a royal pain to solve.
Would be nice, though

Similar Messages

Using Image as alpha layter

I'm reading an Image from a ByteArrayInputStream. This image is purely black-and-white, where the white parts of it should be transparent and the black parts should not. The problem is, the white parts are not transparent.
Is it possible to create an alpha mask from this image, where all white pixels are included in the mask, and then apply this alpha mask onto the original image (which is actually the same image, without being converted to an alpha mask), so only the black pixels are displayed?
Are there any alternative solution if this isn't possible? I'm a bit out of ideas myself at this point :/
(Sorry for the cross-post: http://forum.java.sun.com/thread.jsp?forum=5&thread=485556&tstart=0&trange=15)

This might help.
public static void makeTransparent(BufferedImage bImage, Color color) {
int markColor = color.getRGB() | 0xFF000000;
for (int y=0; y<bImage.getHeight(); y++ ) {
for (int x=0; x<bImage.getWidth(); x++ ) {
int col = bImage.getRGB(x,y);
if ( ( col | 0xFF000000 ) == markColor ) {
bImage.setRGB(x,y, 0x00FFFFFF & col); // Mark the alpha bits as zero - transparent
just do this:
(u have to have some Image image)
int w = image.getWidth(null);
int h = image.getHeight(null);
BufferedImage sBuf = new BufferedImage(w,h, BufferedImage.TYPE_INT_ARGB);
sBuf.getGraphics().drawImage(image, 0,0, null);
makeTransparent(sBuf, new Color(0,0,0)); // for black to transparent
voila.
did it help ?

Problems distributing application using Imaging Source USBCamera and Labview 2010 - Imaging Source sdk required?!

Hi
I've been having trouble with the Imaging Source camera we own when trying to build applications.
Everything works perfectly on the dev PC with appropriate sdk installed.
When I move to building the application (no errors) and deploying it to another PC the activeX control used to interface with the camera does not appear.
Instead I get 'Control could not be loaded' in red text where the control should be.
As far as I can see the activeX component itself is correctly installed on the PC.
If I install the IC Imaging Control sdk the application begins working correctly.
If I uninstall the sdk from the dev PC I get the message about needing a license if I attempt to add the ActiveX control afresh to the program (via the activeX container)
I've posted on the Imaging Source forums and they suggested I try here.
ORIGINAL POST ON IMAGING SOURCE FORUMS:
I'm using a DMK 41BU02 with labview 2010 + IC Labview extension.
The activeX interface for the camera works fine until I deploy it to a PC that does not have the IC Control software installed.
All I see is a 'Control could not be loaded' message where the activeX box should be.
I've installed VB6 redistributable to the client PC so I know the activeX workings should be present on the PC.
If I install IC Control onto the PC the program starts working correctly!
Has anyone compiled software with Labview 2010 successfully?
Is there a trick when creating the application/installer spec?
Thanks in advance
Dan Acres
REPLY
This is correct. I suppose, the LabVIEW setup creating system does not automatically add the used DLL in the "user.lib" directory of your LabVIEW installation path. What does the LabVIEW support say about this, especially in relation to ActiveX controls, that use the Microsoft licensing mechanism?
The dll he mentions is packaged into the support folder in the application spec.
I have also manually copied the dll into the location referenced by the software (Labview 2010\user.lib\xxxx.dll).
This made no difference.
I've got no experience building these kind of thing into applications so I cannot tell if it is caused by a mistake I've made or Labview or if the Imaging Source softwares have done something wrong.
Many thanks in Advance
Dan Acres

Dan,
I think I might have a solution for you and others who need to include the imaging source active-x control in a deployed executable program.
I solved this problem by creating a project that included the dlls and .ocx files required for the Active X Control (these files are in a folder is created once the IC Software is installed C:\Program Files (x86)\Common Files\IC Imaging Control 3.2) and an executable that registers the Active X Control after the main program is installed.
The simple vi that registered the .ocx control for me upon install of my LabVIEW executable was called RegisterActiveXCtrl.vi. All it does is calls the regsvr32 with the path of my .ocx file.
I bundled my main program PCDemo_Stratom.exe and RegICActiveX.exe into an installer and set the option to run RegICActiveX.exe at the end of installation.
The installer must be installed with administrator rights to get the RegICActiveX.exe regsvr32 call to work.
Hope this helps you/others in the future!
Dan A

Flex 4.5 problem using image in Mobile Application (Android)

I have a problem using flex and mobile application...
I have an image with path = "/mnt/sdcard/images/20120202172020990.jpg"
Simply using
<s:Image id="image" source="/mnt/sdcard/images/20120202172020990.jpg" />
Get me an empty image on device, still using it on computer (desktop build / debug) it works fine...
I have enabled every kind of permission in the app.xml
Why I got this error?

I actually solved the problem using:
image.source = File.desktopDirectory.resolvePath(file_name).url;
using instead this one it will work only under windows...
image.source = File.desktopDirectory.resolvePath(file_name).nativePath;
I don't know if it's a bug or whatelse, but I think that the source property of Image passing a filename that start with / (/ is root for unix based OS, android, linux, ios, etc), will take is as a relative path instead of complete path...

Import image as a mask or How to use an alpha mask for transparency mask

Hello,
Basically I need to know how to import 1 image into another image as a mask.
What I'm trying to do:
I have a bunch of large images that put together make 1 very large image (ortho land images). I also have corresponding images, lets refer to them as mask images, that would mask out parts of the large ortho image I want to import it too. These mask images basically look like a PS mask. They are feathered black to white images. I need to be able to import them as a mask.
Is there away, if so how?
Thanks!!!!

If the masked out areas of your image are to be black, see if this approach fits your needs and forget about creating a mask. Instead bring your "mask image" into your ortho land image using your favorite technique. Now change the blend mode of this mask image to mutiply. You should now see land wherever your mask is white, black where the mask is black and transition regions where the mask has a gray value.
Alternatively, you can make an alpha channel if you want by turning off all layers except your "mask image", switching to the channel palette, Ctrl clicking on the rgb channel (to create a selection) and then clicking on the Save Selection as Channel icon at the bottom of the palette. You can then create a luminance mask by going back to the layer palette, selecting the layer you want to mask and click the Add LAyer Mask icon at the bottom of the palette.
Paulo

Animated window mask using imaging lingo-

Has anyone tried animating a window mask at runtime using
imaging lingo? Does it work and is it smooth on at least a
reasonable cross section of machines? I’m thinking something
like the Media Player skins from
http://www.theskinsfactory.com/skinsfactory/
(specifically the Alienware Invader skin) … although frankly,
even on my brand new duo that skin could animate smoother.

I tried it as well once... the file can be downloaded here:
http://www.chunick.com/director/animated_mask.zip
for fun you might want to check out my window with alpha mask
too:
http://www.chunick.com/director/window_alpha_mask.zip
* (Long) Note: the scrnXtra function ScreenToImage() that was
added by Valentin Schmidt to the scrnXtra (and which is used in the
example) has a memory leak in it so don't run the example too long.
The function was added to the scrnXtra using the ScreenToFile()
function as a quick workaround in the C++ code; so ScreenToFile()
could probably be used instead to get the exact same effect without
any memory leak. The ScreenToMember() cannot be used as it has a
flaw because the TrimWhiteSpace() property of a member is enabled
by default and the C++ code does not address this... so any white
pixels on the border of the image will be trimmed and there's no
way to tell which edge of the image they were trimmed from
(AFAIK).

How to separate image from alpha using Quartz Composer?

Cross posting here and in the developers forums... (sorry for any duplication - also posting to the QC mail list)
I'm attempting to construct a Quartz Composer composition that will take an image containing transparency, and — within the structure of that composition — disect the image so that I can perform separate operations on the alpha mask and the full image (i.e. as if the mask did not exist). This would be akin to opening a PNG or TIFF in GraphicConverter, then accessing the images I want via Show Alpha Channel and Remove Alpha Channel.
I'm able to easily access the alpha channel by applying a Color Martix patch to the image, but I can't seem to figure out how to get to the full "unmasked" image.
Can anyone nudge me in the right direction as to how I could remove the alpha mask from an image?
Thanks!!
John

Hi John,
You could use the Core Image Filter to extract the Alpha channel as a mask, and then other filters to use that mask.
In the 'Settings' pane of 'Patch Inspector' for the 'Core Image Filter' place the following Core Image Kernel routine code:
A Core Image kernel routine
Alpha to Mask White Opaque, Black Transparent
kernel vec4 alphaToBlackWhite(sampler image)
vec4 pixelValue = sample(image, samplerCoord(image)); // Get a pixel
          unpremultiply(pixelValue);
          pixelValue.r = pixelValue.a; // Red channel
          pixelValue.g = pixelValue.a; // Green channel
          pixelValue.b = pixelValue.a; // Blue channel
           pixelValue.a = 1.0; // Alpha Channel make Opaque
          return premultiply(pixelValue);
Alternatively you could rewrite the above Kernel, more details on writing Core Image Kernels is in the developer documentation 'Image Tutorial: Writing Kernels'
Hope this helps,
Neil.

Use different "fx-border-image-source" for first tab and remaining tabs

Hi,
I'm using something like this
.tab {
-fx-padding: 0px 5px -2px 5px;
-fx-background-insets: 0 -20 0 0;
-fx-background-color: transparent;
-fx-text-fill: #c4d8de;
-fx-border-image-source: url("images/tab5.png");
-fx-border-image-slice: 20 20 20 20 fill;
-fx-border-image-width: 20 20 20 20;
-fx-border-image-repeat: stretch;
-fx-font-size: 22px;
.tab:selected {
-fx-border-image-source: url("images/tab-selected5.png");
-fx-text-fill: #333333;
     -fx-background-color: red;*/
to customize the tab appearance of a TabPane.
That worked well. But I need to use a different set of images for just the first tab. Does anyone know a way to accomplish that?
Thanks.

How can I "fix up" the first tab of tab panes that are created after I "fixed up" the first tab of the initial tab pane?
My app allows user to create new tab panes at any moment during program execution.Not easy to answer this one.
The best answer would be to use structural pseudoclasses, but (as David points out), they are not yet implemented.
The trick here is how to identify the first tab of each tab pane so that it can be styled separately from the other panes.
Doing the styling without a dynamic lookup is preferrable to using a dynamic lookup (i.e. when the first tab is created give it a specific style, e.g. tab0).
This is how the charts work, where they set style classes based on series of data, e.g. series0, series1 - this allows you to independently style each series of data.
However the chart stuff has all of that built into the implementation, whereas the tabs don't. To achieve that you would likely need to go into the TabSkin code (http://openjdk.java.net/projects/openjfx/) find out where and how it generates the Tab nodes and write a custom tab skin or extension of the existing one which assigns a numeric style class to each new tab in a pane (e.g tab0, tab1, etc). In other words, not particularly easy if you are unfamilar with the tab skin implementation. You could log a javafx jira feature request to have those style classes set on tabs - file it here => http://javafx-jira.kenai.com.
In the meantime a simple alternative is to use the dynamic lookup method in my previous post and a hack such that whenever you add a new tab pane to the scene you do something like the following:
new Timeline(
new KeyFrame(
    Duration.millis(50),
    new EventHandler<ActionEvent>() {
      @Override public void handle(ActionEvent arg0) {
        Node tab = newTabPane.lookup(".tab");
        if (tab != null) tab.getStyleClass().add("first-tab");
).play();The reason for the Timeline is that I don't really know at what stage the css layout pass is executed. I know that when you initially show the stage and then do a lookup, the css pass seems to have already been done and the lookup will work. But for something that is dynamically added or modified after the scene is displayed - I have no idea when the css layout pass occurs, other than it's some time in the future and not at the time that you add the tabPane to the scene. So, the Timeline introduces a short delay to (hopefully) give the css layout pass time to execute and allow the lookup to work (not return null). Not the best or most efficient solution, but should work for you.

Have placed video-when previewed there are sometimes heavy black lines around the image. If I place white rectangles in front, the black is hidden in Author, but when I preview the black lines are still there..How to use white rectangles to mask edges?

I have placed video - when previewed there are sometimes heavy black lines around the image.
If I place white rectangles in front, the black is hidden in Author, but when I preview the black lines are still there..
How to use white rectangles to mask edges?

Here's view in IBA you can see bounding boxes of white rectangles that
form frame to block out black setions..
Here's iBook preview image of same page, NO WHITE FRAME!
WHY?
BOB SPRAGUE

How can I use single source images in text?

I'm using RoboHelp HTML 11, and I want to use single source images of icons in text, so that I have the ability to update the source image, as needed, and all instances of the image are updated automatically. The look is as follows:
I've tried snippets, but they force a paragraph return in the text which I don't want.
I tried inserting the source code for an image into a variable, but when I inserted the variable into the text, I got an error message in the text.
Is there a way to use variables to do this or some other method?
Thanks.

HI there
The one possible drawback to this approach would be if the image size changes. You *MAY* find that saving a different sized image and replacing one would result in perhaps needing to visit each instance of the image and right-clicking and choosing "Reset Size".
I'm not sure why your variable failed with the image. Generally, it's simple to insert an image into the variable. I typically start the process by typing something like "XX" as the variable value. Then in the UDV pod, select the XX and replace it with the desired image.
Cheers... Rick

Seamless repeating pattern - using embedded images with clipping masks

Hi forum,
We struggle a bit creating a seemless repeating pattern that includes embedded images with clipping masks applied.
Our pattern is a 550x550mm black square with various embedded inserted on top. All of these images have Clipping Masks applied to hide unwanted areas.
These images are placed so they create a seamless repeating pattern and this includes areas that overlap the square.
Once we create a swatch from this pattern the repeatable area becomes too big as it includes the unwanted areas hidden by the clipping masks.
Because the repeatable area is now too big it's actually not repeatable.
We tried various things in order to solve this issue. Pathfinder/crop to delete the unwanted areas in the images etc.
None of them worked good enough.
We though of just livetracing the images, but due to the complexity this is not an option.
Do we really have to do our "cutting" in Photoshop or do you know any fancy tricks for issues like this?
Appreaciate your insight
Kenneth

Kenneth,
You may try the destructive deed, with the Clipping Mask selected:
1) In the Transparency palette/panel dropdown list select anything but Normal (Multiply is fine),
2) Object>Flatten Transparency, just keep the defaults including 100% Vector,
3) Shudder.
This should reduce the size to the Bounding Box of the Clipping Path.

Alpha masks

I searched everywhere for a tutorial explaining how to make an image with an aplha mask ( not uniform transparency ) Can you hel me?

Do you mean making an image that is a mask, or making an image using a mask?
To create an image that will be used as a mask, start with a new BufferedImage having an alpha channel (like TYPE_INT_ARGB). Then draw into that image using any of the Java2D primitives, employing transparency either in the color value or an AlphaComposite. Note that the actual color is not important, only the alpha value will be used.
To then use that image as a mask, simply set an alpha composite type of "source in." After that, anything you draw into the image will be constrained by the alpha values of the image.
For example, the following will create an image with a large letter "A" and draw another image into just the pixels occupied by the letter. Since a BufferedImage with an alpha channel is initialized to all transparent, only the visible pixels within the letter will be used to mask the second image.BufferedImage bi = new BufferedImage(100, 100, BufferedImage.TYPE_INT_ARGB);
Graphics2D g = bi.createGraphics();
Font f = new Font("sansserif", Font.BOLD, 140);
g.setFont(f);
g.setColor(Color.black);
g.drawString("A", 0, 100);
g.setComposite(AlphaComposite.SrcIn);
g.drawImage(anotherInterestingImage, 0, 0, null);

Alpha Mask Limit?

Hello, I recently just had a long problem that I could not
figure out. I couldn't get a particular image to take an alpha
blended mask. Since I've been able to successfully do this
countless times prior (through both the timeline and AS), I wasn't
able to find the reason why this particular image wouldn't accept
it. Naturally, I went insane. If I had runtime caching checked, I
would get no masking at all. If I had it unchecked, I would get a
solid mask.
I've narrowed the issue down to one factor and one factor
alone, the physical size of the image. It doesn't matter the file
size of the image being used, rather its dimensions in relation to
the stage. Shrinking the image worked, blowing up any other image
(as a test) makes it stop working again.
My question is this: What's the largest size image we can
alpha mask? Mine was 1608x259(original) and I was trying to scroll
it across a 705x504 size stage. I didn't keep sizing down until it
worked, rather I just cut it up at its current size, so I don't
know yet this limit, but it's there.

300thnickattempt wrote:
> Hello, I recently just had a long problem that I could
not figure out. I
> couldn't get a particular image to take an alpha blended
mask. Since I've been
> able to successfully do this countless times prior
(through both the timeline
> and AS), I wasn't able to find the reason why this
particular image wouldn't
> accept it. Naturally, I went insane. If I had runtime
caching checked, I would
> get no masking at all. If I had it unchecked, I would
get a solid mask.
>
> I've narrowed the issue down to one factor and one
factor alone, the physical
> size of the image. It doesn't matter the file size of
the image being used,
> rather its dimensions in relation to the stage.
Shrinking the image worked,
> blowing up any other image (as a test) makes it stop
working again.
>
> My question is this: What's the largest size image we
can alpha mask? Mine was
> 1608x259(original) and I was trying to scroll it across
a 705x504 size stage. I
> didn't keep sizing down until it worked, rather I just
cut it up at its current
> size, so I don't know yet this limit, but it's there.
>
I could not reproduce the problem. For test I imported 2000
by 700 pixels image.
Set caching bitmap and mask using setMask action, animated
the mask, animate the image
and than the both in the same time.
Variety of combinations, each time it work w/o problem.
My guess is the file get somehow corrupted. FLA are very
delicate and its common to
experience unusually behaviors for no apparent reason.
Try to do the same in new file see if happens again. Try to
adjust the dimensions, make
the image slightly bigger or smaller. Remove the image from
stage, drag it from library
to stage again and convert it to an instance movie clip.
Sometimes the link between
library content and stage get messed up.
Regards
Urami
Happy New Year guys - all the best there is in the 2006 :)
<urami>
http://www.Flashfugitive.com
</urami>
<web junk free>
http://www.firefox.com
</web junk free>

Problem using images in a button skin

So, I have a weird problem. I need to be able to see a button in design mode. My button is a custom component (a ButtonBase) with a skin. So, I left is a tiny bit of the default skin so that you can see what it looked like in design mode. Now, the paths to the images that will be used for the various states of the button are binded to strings passed from another subcomponent. There was no other way to get what we needed done, so it stays messy. Anyway, I don't get things to display correctly. Here's the code, and after an image for what I'm getting:
[CODE]
<s:SparkSkin xmlns:fx="http://ns.adobe.com/mxml/2009"
                         xmlns:s="library://ns.adobe.com/flex/spark"
                         xmlns:fb="http://ns.adobe.com/flashbuilder/2009"
                         xmlns:mx="library://ns.adobe.com/flex/mx"
                         minWidth="21" minHeight="21" alpha.disabled="0.5"
                         preinitialize="initImageVars()" creationComplete="init()">
    
    <fx:Metadata>
        <![CDATA[
        [HostComponent("customComponents.ToggleButtonCustom")]
        ]]>
    </fx:Metadata>
    
    <s:states>
        <s:State name="up" />
        <s:State name="down" />
        <s:State name="selected" />
        <s:State name="disabled" />
        <s:State name="over" />
    </s:states>
    <fx:Script>
        <![CDATA[
            import mx.controls.Alert;
            /* First thing's first: we need to initialize our global and accessable components. */
            [Bindable]
            public var upButtonPath:String = "";
            [Bindable]
            public var downButtonPath:String = "";
            [Bindable]
            public var activeButtonPath:String = "";
            [Bindable]
            public var disabledButtonPath:String = "";
            [Bindable]
            protected var label:String = "";
            private function initImageVars():void
                upButtonPath = "assets/" + this.hostComponent.upButtonPath;
                downButtonPath = "assets/" + this.hostComponent.downButtonPath;
                activeButtonPath = "assets/" + this.hostComponent.activeButtonPath;
                disabledButtonPath = "assets/" + this.hostComponent.disabledButtonPath;
            private function init():void
                /* Give values to public variables */
                //upButtonPath = "Up.png";
                //downButtonPath = "Down.png";
                //activeButtonPath = "Active.png";
                label = this.hostComponent.label;
            /* Define the skin elements that should not be colorized. For button, the graphics are colorized but the label is not. */
            static private const exclusions:Array = ["labelDisplay"];
            override public function get colorizeExclusions():Array {return exclusions;}
            override protected function initializationComplete():void
                useChromeColor = true;
                super.initializationComplete();
            private var cornerRadius:Number = 2;
        ]]>
    </fx:Script>
    
    <s:Rect id="fill" left="1" right="1" top="1" bottom="1" radiusX="2">
        <s:fill>
            <s:LinearGradient rotation="90">
                <s:GradientEntry color="0xFFFFFF"
                                                 color.down="0xAAAAAA"
                                                 color.selected="0xBBBDBD"
                                                 alpha="0.85" />
                <s:GradientEntry color="0xD8D8D8"
                                                 color.down="0x929496"
                                                 color.selected="0x9FA0A1"
                                                 alpha="0.85" />
            </s:LinearGradient>
        </s:fill>
    </s:Rect>
    <mx:ViewStack id="buttonImage" x="0" y="0"
                                selectedIndex="0" selectedIndex.down="1" selectedIndex.selected="2"    selectedIndex.disabled="3">
        <s:NavigatorContent id="upState">
            <mx:Image source="{upButtonPath}" width="500" height="500" />
        </s:NavigatorContent>
        <s:NavigatorContent id="downState">
            <mx:Image source="{downButtonPath}" />
        </s:NavigatorContent>
        <s:NavigatorContent id="activeSate">
            <mx:Image source="{activeButtonPath}" />
        </s:NavigatorContent>
        <s:NavigatorContent id="disabledState">
            <mx:Image source="{disabledButtonPath}" />
        </s:NavigatorContent>
    </mx:ViewStack>
    <s:Label id="labelDisplay" text="{label}"
                     textAlign="center"
                     verticalAlign="middle"
                     maxDisplayedLines="3"
                     horizontalCenter="0" verticalCenter="1"
                     left="10" right="10" top="2" bottom="2">
    </s:Label>
</s:SparkSkin>
[/CODE]
That gives me this image:
And if I use the following code to resize it, I get this:
[CODE]
/* From the component, called from init(), which is called on creationComplete */
private function resizeEverything():void
var dummyImage:Image = new Image();
var pathToDummy:String = "assets" + upButtonPath;
dummyImage.source = pathToDummy;
thisBtn.height = dummyImage.height;
thisBtn.width = dummyImage.width;
dummyImage = null;
[/CODE]
If you have any idea what to do with this, let me know. I set the buttons to the exact height and width of the images I'm using when I first call them. I don't know how else to manage this. I want the faded white to be an exact mirror of the image, so that designers (not me) can move them around on design mode. Thank you!

Figured it out. The images I was given had large amounts of padding, and when I tried to auto-resize the buttons, Flash could never grab the image and therefore couldn't size off it, so the size of the component was forced to 0,0. It's all good now.

Getting rid of black fringe when using anti-alised alpha channel

I want to take images rendered in Lightwave 9.3 into Photoshop CS2 for compositing. If:
1. I render objects with antialiased edges over a black background, and then
2. save out the alpha channel for use as a layer mask in Photoshop
I get a black fringe around the objects resulting from the semitransparent pixels of the antialiased edges revealing the black background in the base image.
How should I avoid this problem to get clean antialiased edges?
Thanks.

One of the few uses of the Image>Adjust>Brightness/Contrast is choking or spreading Alpha channels. Go to the channels palette and play around with the Alpha to see what happens. Bear in mind that you may want to load the channel as a selection (or not), invert the selection (or not), Modify the selection to expand or contract... (or not), etc, etc.
The point of all the "or nots" is that there is more than one way to attack this. Your willingness to experiment will determine whether you succeed (or not).
Here's a before and after:
http://img369.imageshack.us/img369/103/alphaon3.jpg

Animate Particles Using Image Source Alpha Mask - Outline

Similar Messages

Maybe you are looking for