Canvas QML Type

Provides a 2D canvas item enabling drawing via JavaScript More...

Import Statement: import QtQuick 2.7
Since: Qt 5.0


Inherited By:





Detailed Description

The Canvas item allows drawing of straight and curved lines, simple and complex shapes, graphs, and referenced graphic images. It can also add text, colors, shadows, gradients, and patterns, and do low level pixel operations. The Canvas output may be saved as an image file or serialized to a URL.

Rendering to the Canvas is done using a Context2D object, usually as a result of the paint signal.

To define a drawing area in the Canvas item set the width and height properties. For example, the following code creates a Canvas item which has a drawing area with a height of 100 pixels and width of 200 pixels:

import QtQuick 2.0
Canvas {
    id: mycanvas
    width: 100
    height: 200
    onPaint: {
        var ctx = getContext("2d");
        ctx.fillStyle = Qt.rgba(1, 0, 0, 1);
        ctx.fillRect(0, 0, width, height);

Currently the Canvas item only supports the two-dimensional rendering context.

Threaded Rendering and Render Target

The Canvas item supports two render targets: Canvas.Image and Canvas.FramebufferObject.

The Canvas.Image render target is a QImage object. This render target supports background thread rendering, allowing complex or long running painting to be executed without blocking the UI. This is the only render target that is supported by all Qt Quick backends.

The Canvas.FramebufferObject render target utilizes OpenGL hardware acceleration rather than rendering into system memory, which in many cases results in faster rendering. Canvas.FramebufferObject relies on the OpenGL extensions GL_EXT_framebuffer_multisample and GL_EXT_framebuffer_blit for antialiasing. It will also use more graphics memory when rendering strategy is anything other than Canvas.Cooperative. Framebuffer objects may not be available with Qt Quick backends other than OpenGL.

The default render target is Canvas.Image and the default renderStrategy is Canvas.Immediate.

Pixel Operations

All HTML5 2D context pixel operations are supported. In order to ensure improved pixel reading/writing performance the Canvas.Image render target should be chosen. The Canvas.FramebufferObject render target requires the pixel data to be exchanged between the system memory and the graphic card, which is significantly more expensive. Rendering may also be synchronized with the V-sync signal (to avoid screen tearing) which will further impact pixel operations with Canvas.FrambufferObject render target.

Tips for Porting Existing HTML5 Canvas Applications

Although the Canvas item provides an HTML5-like API, HTML5 canvas applications need to be modified to run in the Canvas item:

  • Replace all DOM API calls with QML property bindings or Canvas item methods.
  • Replace all HTML event handlers with the MouseArea item.
  • Change setInterval/setTimeout function calls with the Timer item or the use of requestAnimationFrame().
  • Place painting code into the onPaint handler and trigger painting by calling the markDirty() or requestPaint() methods.
  • To draw images, load them by calling the Canvas's loadImage() method and then request to paint them in the onImageLoaded handler.

Starting Qt 5.4, the Canvas is a texture provider and can be used directly in ShaderEffects and other classes that consume texture providers.

Note: In general large canvases, frequent updates, and animation should be avoided with the Canvas.Image render target. This is because with accelerated graphics APIs each update will lead to a texture upload. Also, if possible, prefer QQuickPaintedItem and implement drawing in C++ via QPainter instead of the more expensive and likely less performing JavaScript and Context2D approach.

See also Context2D and QQuickPaintedItem.

Property Documentation

available : bool

Indicates when Canvas is able to provide a drawing context to operate on.

canvasSize : size

Holds the logical canvas size that the context paints on.

By default, the canvas size is the same size as the current canvas item size.

By setting the canvasSize, tileSize and canvasWindow, the Canvas item can act as a large virtual canvas with many separately rendered tile rectangles. Only those tiles within the current canvas window are painted by the Canvas render engine.

See also tileSize and canvasWindow.

context : object

Holds the active drawing context.

If the canvas is ready and there has been a successful call to getContext() or the contextType property has been set with a supported context type, this property will contain the current drawing context, otherwise null.

contextType : string

The type of drawing context to use.

This property is set to the name of the active context type.

If set explicitly the canvas will attempt to create a context of the named type after becoming available.

The type name is the same as used in the getContext() call, for the 2d canvas the value will be "2d".

See also getContext() and available.

renderStrategy : enumeration

Holds the current canvas rendering strategy.

  • Canvas.Immediate - context will perform graphics commands immediately in the main UI thread.
  • Canvas.Threaded - context will defer graphics commands to a private rendering thread.
  • Canvas.Cooperative - context will defer graphics commands to the applications global render thread.

This hint is supplied along with renderTarget to the graphics context to determine the method of rendering. A renderStrategy, renderTarget or a combination may not be supported by a graphics context, in which case the context will choose appropriate options and Canvas will signal the change to the properties.

Configuration or runtime tests may cause the QML Scene Graph to render in the GUI thread. Selecting Canvas.Cooperative, does not guarantee rendering will occur on a thread separate from the GUI thread.

The default value is Canvas.Immediate.

See also renderTarget.

renderTarget : enumeration

Holds the current canvas render target.

  • Canvas.Image - render to an in memory image buffer.
  • Canvas.FramebufferObject - render to an OpenGL frame buffer

This hint is supplied along with renderStrategy to the graphics context to determine the method of rendering. A renderStrategy, renderTarget or a combination may not be supported by a graphics context, in which case the context will choose appropriate options and Canvas will signal the change to the properties.

The default render target is Canvas.Image.

Signal Documentation


This signal is emitted when an image has been loaded.

The corresponding handler is onImageLoaded.

See also loadImage().

paint(rect region)

This signal is emitted when the region needs to be rendered. If a context is active it can be referenced from the context property.

This signal can be triggered by markdirty(), requestPaint() or by changing the current canvas window.

The corresponding handler is onPaint.


This signal is emitted after all context painting commands are executed and the Canvas has been rendered.

The corresponding handler is onPainted.

Method Documentation

cancelRequestAnimationFrame(long handle)

This function will cancel the animation callback referenced by handle.

object getContext(string contextId, ... args)

Returns a drawing context, or null if no context is available.

The contextId parameter names the required context. The Canvas item will return a context that implements the required drawing mode. After the first call to getContext, any subsequent call to getContext with the same contextId will return the same context object.

If the context type is not supported or the canvas has previously been requested to provide a different and incompatible context type, null will be returned.

Canvas only supports a 2d context.

isImageError(url image)

Returns true if the image failed to load.

See also loadImage().

isImageLoaded(url image)

Returns true if the image is successfully loaded and ready to use.

See also loadImage().

isImageLoading(url image)

Returns true if the image is currently loading.

See also loadImage().

loadImage(url image)

Loads the given image asynchronously.

When the image is ready, imageLoaded will be emitted. The loaded image can be unloaded by the unloadImage() method.

Note: Only loaded images can be painted on the Canvas item.

See also unloadImage, imageLoaded, isImageLoaded(), Context2D::createImageData(), and Context2D::drawImage().

markDirty(rect area)

Mark the given area as dirty, so that when this area is visible the canvas renderer will redraw it. This will trigger the paint signal.

See also paint and requestPaint().

long requestAnimationFrame(callback)

This function schedules callback to be invoked before composing the Qt Quick scene.


Request the entire visible region be re-drawn.

See also markDirty().

bool save(string filename)

Save the current canvas content into an image file filename. The saved image format is automatically decided by the filename's suffix.

Note: calling this method will force painting the whole canvas, not just the current canvas visible window.

See also canvasWindow, canvasSize, and toDataURL().

string toDataURL(string mimeType)

Returns a data URL for the image in the canvas.

The default mimeType is "image/png".

See also save().

unloadImage(url image)

Unloads the image.

Once an image is unloaded it cannot be painted by the canvas context unless it is loaded again.

See also loadImage(), imageLoaded, isImageLoaded(), Context2D::createImageData(), and Context2D::drawImage.

© 2017 The Qt Company Ltd. Documentation contributions included herein are the copyrights of their respective owners. The documentation provided herein is licensed under the terms of the GNU Free Documentation License version 1.3 as published by the Free Software Foundation. Qt and respective logos are trademarks of The Qt Company Ltd. in Finland and/or other countries worldwide. All other trademarks are property of their respective owners.