375e894038
* Make Frame Processors an extra subspec * Update VisionCamera.podspec * Make optional * Make VisionCamera compile without Skia * Fix * Add skia again * Update VisionCamera.podspec * Make VisionCamera build without Frame Processors * Rename error to `system/frame-processors-unavailable` * Fix Frame Processor returning early * Remove `preset`, FP partial rewrite * Only warn on frame drop * Fix wrong queue * fix: Run on CameraQueue again * Update CameraView.swift * fix: Activate audio session asynchronously on audio queue * Update CameraView+RecordVideo.swift * Update PreviewView.h * Cleanups * Cleanup * fix cast * feat: Add LiDAR Depth Camera support * Upgrade Ruby * Add vector icons type * Update Gemfile.lock * fix: Stop queues on deinit * Also load `builtInTrueDepthCamera` * Update CameraViewManager.swift * Update SkImageHelpers.mm * Extract FrameProcessorCallback to FrameProcessor Holds more context now :) * Rename to .m * fix: Add `RCTLog` import * Create SkiaFrameProcessor * Update CameraBridge.h * Call Frame Processor * Fix defines * fix: Allow deleting callback funcs * fix Skia build * batch * Just call `setSkiaFrameProcessor` * Rewrite in Swift * Pass `SkiaRenderer` * Fix Import * Move `PreviewView` to Swift * Fix Layer * Set Skia Canvas to Frame Host Object * Make `DrawableFrameHostObject` subclass * Fix TS types * Use same MTLDevice and apply scale * Make getter * Extract `setTorch` and `Preview` * fix: Fix nil metal device * Don't wait for session stop in deinit * Use main pixel ratio * Use unique_ptr for Render Contexts * fix: Fix SkiaPreviewDisplayLink broken after deinit * inline `getTextureCache` * Update CameraPage.tsx * chore: Format iOS * perf: Allow MTLLayer to be optimized for only frame buffers * Add RN Video types * fix: Fix Frame Processors if guard * Find nodeModules recursively * Create `Frame.isDrawable` * Add `cocoapods-check` dependency |
||
---|---|---|
.github | ||
.vscode | ||
android | ||
cpp | ||
docs | ||
example | ||
ios | ||
scripts | ||
src | ||
.eslintrc.js | ||
.gitignore | ||
app.plugin.js | ||
CODE_OF_CONDUCT.md | ||
CONTRIBUTING.md | ||
LICENSE | ||
package.json | ||
README.md | ||
tsconfig.json | ||
VisionCamera.podspec | ||
yarn.lock |
Vision Camera
‼️‼️‼️‼️‼️ ✨ VisionCamera V3 ‼️‼️‼️‼️‼️
See this discussion for the latest upcoming version of VisionCamera
Documentation
Features
- Photo, Video and Snapshot capture
- Customizable devices and multi-cameras (smoothly zoom out to "fish-eye" camera)
- Customizable FPS
- Frame Processors (JS worklets to run QR-Code scanning, facial recognition, AI object detection, realtime video chats, ...)
- Smooth zooming (Reanimated)
- Fast pause and resume
- HDR & Night modes
See the example app
Example
function App() {
const devices = useCameraDevices('wide-angle-camera')
const device = devices.back
if (device == null) return <LoadingView />
return (
<Camera
style={StyleSheet.absoluteFill}
device={device}
isActive={true}
/>
)
}
Adopting at scale
VisionCamera is provided as is, I work on it in my free time.
If you're integrating VisionCamera in a production app, consider funding this project and contact me to receive premium enterprise support, help with issues, prioritize bugfixes, request features, help at integrating VisionCamera and/or Frame Processors, and more.