tencent cloud

$0 14-Day TrialExperience EdgeOne for acceleration and security protection!

Feedback

Tencent Effect SDK

Configuring Segmentation

Last updated: 2024-05-08 16:34:16
The segmentation module needs to be enabled during initialization to implement virtual backgrounds. For more information, see Custom Stream or Image and Built-in Camera.


Setting the Background

The SDK allows you to blur the background or set an image as the background. You can pass in keying parameters during initialization.
const config = {
module: {
beautify: true, // Whether to enable the effect module, which offers beautification and makeup effects as well as stickers
segmentation: true // Whether to enable the segmentation module, which allows you to change the background
segmentationLevel: 0 // Switch background segmentation models supported since Versions 1.0.19
},
auth: authData, // The authentication information
input: stream, // The input stream
beautify: { // The effect parameters for initialization (optional)
whiten: 0.1,
dermabrasion: 0.3,
eye: 0.2,
chin: 0,
lift: 0.1,
shave: 0.2
},
background: {
type: 'blur' // Blur the background
}
}
const sdk = new ArSdk(
// Pass in a config object to initialize the SDK
config
)
You can also dynamically change the background.
sdk.setBackground({
type: 'image', // The background image
src: 'https://webar-static.tencent-cloud.com/assets/background/1.jpg'
})

Using Transparent Backgrounds

The SDK supports transparent backgrounds on some browsers.
sdk.setBackground({
type: 'transparent'
})
Note:
Keep the following in mind:
Segmentation is supported on both mobile and desktop browsers.
Because WebRTC does not support alpha channels, you can only use transparent backgrounds locally. Background transparency will not work after publishing.
Background transparency is supported on desktop Chrome and Firefox, but not supported on desktop or mobile Safari.
Starting from version 1.0.19 and above, switching background segmentation models is supported, with parameters: 0, 1, 2.
Level 0 has the best performance but relatively average segmentation results.
Level 1 has a moderate performance and effect.
Level 2 has the best segmentation results and the longest inference time, suitable for applications with high requirements for segmentation effects and low performance requirements.


Catalog

In The Article

Configuring Segmentation

Contact Us

Contact our sales team or business advisors to help your business.

Technical Support

Open a ticket if you're looking for further assistance. Our Ticket is 7x24 avaliable.

7x24 Phone Support
Hong Kong, China
+852 800 906 020 (Toll Free)
United States
+1 844 606 0804 (Toll Free)
United Kingdom
+44 808 196 4551 (Toll Free)
Canada
+1 888 605 7930 (Toll Free)
Australia
+61 1300 986 386 (Toll Free)
EdgeOne hotline
+852 300 80699
More local hotlines coming soon