I have a rectangle on a canvas that the user can resize, move and so on to make a selection.
I also have an image the size of the screen behind the canvas (basically a screenshot).
I'd like to translate the selection (the rectangle) in the canvas to a 1:1 selection in the image (I want the image directly behind the rectangle) given I have the rectangle's Canvas.Top, Canvas.Left, Width, Height.
<Grid Name="MainGrid" SnapsToDevicePixels="False" HorizontalAlignment="Stretch" VerticalAlignment="Stretch">
<Image x:Name="MainImage" Stretch="None" RenderOptions.BitmapScalingMode="HighQuality"/>
<Border Background="Black" Opacity="0.4" HorizontalAlignment="Stretch" VerticalAlignment="Stretch"/>
<Canvas Name="MainCanvas" Width="{Binding Source={x:Static SystemParameters.PrimaryScreenWidth}}" Height="{Binding Source={x:Static SystemParameters.PrimaryScreenHeight}}" Background="Transparent">
<ContentControl Name="SelectionRect" />
</ContentControl>
</Canvas>
</Grid>
I tried doing this: (MainImage is the image under the canvas)
Rect rect = new Rect(Canvas.GetLeft(SelectionRect), Canvas.GetTop(SelectionRect), SelectionRect.Width, SelectionRect.Height);
Rect from_rect = SelectionRect.TransformToVisual(this).TransformBounds(rect);
BitmapSource cropped_bitmap = new CroppedBitmap(MainImage.Source as BitmapSource,
new Int32Rect((int)from_rect.X, (int)from_rect.Y, (int)from_rect.Width, (int)from_rect.Height));
SelectionRectImageSource = cropped_bitmap;
But the image I get (SelectionRectImageSource) is a moved aside version of the actual pixels behind the selection rectangle.
So basically, I don't understand how these transformations work and how I should use them if at all.
Example:
Thanks a lot!
Dolev.
Looks like you need to correct for the DPI difference between the image (usually 72dpi) and the presentation source (usually 96dpi). Additionally, your first Rect should not be offset by Canvas.Left and Canvas.Top; TransformToVisual will take care of the relative offset for you.
var source = (BitmapSource)MainImage.Source;
var selectionRect = new Rect(SelectionRect.RenderSize);
var sourceRect = SelectionRect.TransformToVisual(MainImage)
.TransformBounds(selectionRect);
var xMultiplier = source.PixelWidth / MainImage.ActualWidth;
var yMultiplier = source.PixelHeight / MainImage.ActualHeight;
sourceRect.Scale(xMultiplier, yMultiplier);
var croppedBitmap = new CroppedBitmap(
source,
new Int32Rect(
(int)sourceRect.X,
(int)sourceRect.Y,
(int)sourceRect.Width,
(int)sourceRect.Height));
SelectionRectImageSource= croppedBitmap;
Depending on where this code resides, you may also need to transform the selection rectangle to MainImage instead of this (as I did).
Also, in case MainImage.Source is smaller than the actual MainImage control, you should probably set the horizontal and vertical alignments of MainImage to Left and Top, respectively, less your translated rectangle end up outside the bounds of the source image. You'll need to clamp the selection rectangle to the dimensions of MainImage too.
Related
NEW INFORMATION!!!
Meanwhile i have found a solution but there is a new problem.
The solution is to set the margin of the canvas in code-behind to a new object of type Thickness with top and left 1 or 2.
But the canvas is lying on a tabcontrol.
When i switch between tabs or make a mousedown on the canvas the margin is lost.
I'm working with VS2015 on a WPF-application and have a very curious problem.
I got in one of my WPF windows a canvas as parent for some child elements.
One of these elements is a rectangle which shall show the user the size of a DIN-A4 page.
It is added in code-behind to the children collection of the canvas.
Normally i would place that rectangle at position (0,0).
But because of some problems i have to trick and set the position to (-1,-1) like that:
public static System.Windows.Shapes.Rectangle GetRectangle(double top, double left, double width, double height)
{
var rectangle = new System.Windows.Shapes.Rectangle();
System.Windows.Controls.Canvas.SetLeft(rectangle, -1);
System.Windows.Controls.Canvas.SetTop(rectangle, -1);
rectangle.Width = Math.Round(GetSizeInPoint(width)) + 2;
rectangle.Height = Math.Round(GetSizeInPoint(height));
rectangle.StrokeThickness = 1;
rectangle.Stroke = System.Windows.Media.Brushes.Red;
return rectangle;
}
But the result of it is that just a small part of the top and left border of the rectangle can be seen.
Do i have a chance to "move" the canvas so that the rectangle is displayed completely?
Hereby another important problematic point is that the Grid named "grdProtocolDesigner" can be serialized to XML and saved in the database.
So a complete restructuring would be a big problem.
Here the relevant part of my XAML including the canvas:
<ContentPresenter x:Name="protocolContainer"
Grid.Row="1"
Grid.Column="0">
<ContentPresenter.Content>
<Grid x:Name="grdProtocolDesigner"
Grid.Row="1"
Grid.Column="0">
<ScrollViewer>
<!--Important! The background of this panel has to be "Transparent" so that drag'n'drop-events can work.
Otherwise the events are not fired.-->
<Canvas x:Name="protocolDesignerPanel"
AllowDrop="True"
Visibility="{Binding DesignerPanelsVisibility}"
Width="2200" Height="4000"
MouseEnter="designerpanel_MouseEnter"
MouseLeave="designerpanel_MouseLeave"
MouseDown="designerpanel_MouseDown"
MouseMove="designerpanel_MouseMove"
MouseUp="designerPanel_MouseUp">
<Canvas.RenderTransform>
<ScaleTransform x:Name="scaleProtocolLayout"/>
</Canvas.RenderTransform>
<Canvas.Background>
<ImageBrush ImageSource="../../pictures/CanvasBackground.png"
TileMode="Tile"
Stretch="None"
Viewport="0, 0, 10, 10"
ViewportUnits="Absolute" />
</Canvas.Background>
</Canvas>
</ScrollViewer>
</Grid>
</ContentPresenter.Content>
</ContentPresenter>
This my first post on stack overflow, I hope I get it right. I am using the ZoomControl from WPF Extensions to display an image with pan and zoom support:
<DockPanel Grid.Row="1" x:Name="canvas">
<Controls:ZoomControl x:Name="zoomControl">
<Canvas x:Name="canvas">
<Image x:Name="imageControl" Stretch="None" />
</Canvas>
</Controls:ZoomControl>
</DockPanel>
When the user selects an image with a bowse dialog, I load that image like so:
bmp = new BitmapImage(new Uri(fileName));
this.imageControul.Source = bmp;
I would like to added rectangles\adorners to specific locations (pixel coordinates) on the image the user loaded based on some image processing.
var r = new Rectangle();
r.StrokeThickness = 5;
r.Stroke = Brushes.Black;
r.Fill = Brushes.Transparent;
r.Width = width;
r.Height = height;
Canvas.SetLeft(r, y);
Canvas.SetTop(r, x);
canvas.Children.Add(r);
However, the rectangles are not placed in the expected locations? Wrong scale and location.
Thanks,
John
I expect the problem is that your Canvas is expanding to fill the space rather than being locked to the rectangle. Have a look with a tool like Snoop and see what the bounding boxes of the two are.
You might be able to fix it with Horizontal and VercticalAlignment on the canvas, set them to anything other than Stretch.
If that doesn't work restructure it like this
<ZoomBox>
<Grid>
<Image/>
<Canvas/>
</Grid>
</ZoomBox>
So the Image and the canvas are grouped by the parent Grid which is being transformed.
I have a grid that contains an image.
<Grid Name="grid1">
<Image Name="image1" Stretch="None" Source="C:\Users\User\Desktop\image.jpg"/>
</Grid>
If the size of image was greater than the size of the grid, I want to scale it manually by render transform to fit the grid. I don't want to use Stretch="Fill" because I need the scale factor.
Is there any way to detect the situation that an UIElement goes out of view?
I need your help.
You could simply set the Stretch property to Uniform (or perhaps Fill) and calculate the scaling factor from the ActualWidth of the Image and the Width of the ImageSource, whenever you need it. The sample below does the calculation in a SizeChanged handler, but it could be anywhere else.
<Image Name="image1" Stretch="Uniform" Source="C:\Users\User\Desktop\image.jpg"
SizeChanged="ImageSizeChanged"/>
The calculation looks like this:
private void ImageSizeChanged(object sender, SizeChangedEventArgs e)
{
var scale = image1.ActualWidth / image1.Source.Width;
}
As Uniform is the default value of the Stretch property, you wouldn't have to set it at all.
I'm not sure why you would want to rescale your image manually when WPF can do it for you...
Instead of putting the image straight into the grid, use a Viewbox control:
<Grid Name="grid1">
<Viewbox>
<Image Name="image1" Stretch="None" Source="C:\Users\User\Desktop\image.jpg"/>
</Viewbox>
</Grid>
The Viewbox will automatically scale the picture to fit inside the grid...
I would like to zoom an image in WPF and that the image visual render be inside a constrained sized item control.
For example:
<Canvas x:Name="m_canvas" MaxWidth="300" MaxHeight="300" >
<Image Source="..."
Width="300"
Height="300" />
</Canvas>
The zoom code:
var matrix = ((MatrixTransform)m_image.RenderTransform).Matrix;
var center = new Point(m_image.ActualWidth / 2, m_image.ActualHeight / 2);
center = matrix.Transform(center);
matrix.ScaleAt(delta.Scale.X, delta.Scale.Y, center.X, center.Y);
((MatrixTransform)m_image.RenderTransform).Matrix = matrix;
The problem is that when I'm zooming the image render size go larger that the canvas limit (300x300). I would like if the image can zoom only in the canvas.
I don't want to limit the max zoom, I want that if the render size of the image is larger that the canvas, it's stay inside. I don't want that it overlap the canvas
You could clip to the bounds of the Canvas:
<Canvas ClipToBounds="True" ...>
But I don't understand why you're using a Canvas in the first place. It's likely that there's a much nicer way to approach your particular problem without the need for hard-coded widths and heights and without any Canvas at all.
I am currently designing a layout app in Silverlight and have a Canvas inside of a Viewbox. I add shapes to the canvas and they display properly, when I resize the viewbox to zoom in at 2x the height and width, everything still draws properly.
The problem comes when I try to zoom at a factor of 4 or greater or at 0.5 (zoomed out).
Update: The horizontal lines are still there, they are just not drawing. Interaction between the the other shapes and the disappearing ones is still present
When I do this, any horizontal lines do not redraw, but any other shapes, vertical lines of other, still redraw fine. The objects are still children of the canvas and their visibilities are all set to visible.
What is happening?
Update
Very Simple XAML:
<ScrollViewer x:Name="scrollViewer"
Padding="0"
ScrollViewer.VerticalScrollBarVisibility="Auto"
ScrollViewer.HorizontalScrollBarVisibility="Auto"
IsTabStop="False"
Background="Beige">
<Viewbox x:Name="viewBox" Stretch="UniformToFill">
<Canvas x:Name="designCanvas"
Background="{Binding ElementName=mainControl, Path=Background, Mode=TwoWay}">
</Canvas>
</Viewbox>
</ScrollViewer>
Here is how I add the shapes:
Rectangle horGuide = new Rectangle()
{
Tag = "horGuide",
Fill = new SolidColorBrush(Colors.Cyan),
Height = 0.5,
Width = designCanvas.canvActualWidth*16,
};
int h = designCanvas.horOffset;
int v = designCanvas.vertOffset;
double d = e.GetPosition(sideRule).Y;
designCanvas.Children.Add(horGuide);
Canvas.SetTop(horGuide, ((d+v )/ designCanvas.zoomFactor));
Canvas.SetLeft(horGuide, 0 - h);
To Zoom in:
viewBox.Width *= 2;
viewBox.Height *= 2;
Why not use the ScaleTransform class to zoom in and out.