视频中的AVfoundation模糊背景

Question

在我的应用程序中,我修复了1280 x 720的合成渲染大小.因此,如果要导入任何肖像视频,那么我必须在中心显示带有填充和视频宽高比的模糊背景.像这样:

https://www.youtube.com/watch?v=yCOrqUA0ws4

我实现了使用两种视频AVMtableComposition,但我不知道如何模糊特定的背景音轨.我在我的代码中做了以下事情:

self.composition = [AVMutableComposition composition];
AVAsset *firstAsset = [AVAsset assetWithURL:[NSURL fileURLWithPath:[[NSBundle mainBundle] pathForResource:@"ScreenFlow_Blend" ofType:@"mp4"]]];


[self addAsset:firstAsset toComposition:self.composition withTrackID:1];
[self addAsset:firstAsset toComposition:self.composition withTrackID:2];
//  [self addAsset:ThirdAsset toComposition:self.composition withTrackID:3];

AVAssetTrack *backVideoTrack = [firstAsset tracksWithMediaType:AVMediaTypeVideo][0];;

self.videoComposition = [AVMutableVideoComposition videoComposition];
self.videoComposition.renderSize = CGSizeMake(1280, 720);
self.videoComposition.frameDuration = CMTimeMake(1, 30);

AVMutableVideoCompositionInstruction *instruction = [AVMutableVideoCompositionInstruction videoCompositionInstruction];
instruction.timeRange = [backVideoTrack timeRange];

CGFloat scale = 1280/backVideoTrack.naturalSize.width;
CGAffineTransform t = CGAffineTransformMakeScale(scale, scale);
t = CGAffineTransformTranslate(t, 0, -backVideoTrack.naturalSize.height/2 + self.videoComposition.renderSize.height/2);

AVMutableVideoCompositionLayerInstruction *frontLayerInstruction = [AVMutableVideoCompositionLayerInstruction videoCompositionLayerInstruction];
frontLayerInstruction.trackID = 1;
[frontLayerInstruction setTransform:t atTime:kCMTimeZero];

CGFloat scaleSmall = 720/backVideoTrack.naturalSize.height;

CGAffineTransform  translate = CGAffineTransformMakeTranslation(self.videoComposition.renderSize.width/2 - ((backVideoTrack.naturalSize.width/2)*scaleSmall),0);

CGAffineTransform  scaleTransform = CGAffineTransformMakeScale(scaleSmall,scaleSmall);

CGAffineTransform finalTransform = CGAffineTransformConcat(scaleTransform, translate);


CGAffineTransform t1 = CGAffineTransformMakeScale(scaleSmall,scaleSmall);
t1 = CGAffineTransformTranslate(t1,1280, 0);

AVMutableVideoCompositionLayerInstruction *backLayerInstruction = [AVMutableVideoCompositionLayerInstruction videoCompositionLayerInstruction];
backLayerInstruction.trackID = 2;
[backLayerInstruction setTransform:finalTransform atTime:kCMTimeZero];


//    AVMutableVideoCompositionLayerInstruction *maskLayerInstruction = [AVMutableVideoCompositionLayerInstruction videoCompositionLayerInstruction];
//    maskLayerInstruction.trackID = 3;
//    [maskLayerInstruction setTransform:t atTime:kCMTimeZero];


instruction.layerInstructions = @[backLayerInstruction,frontLayerInstruction];

self.videoComposition.instructions = @[ instruction ];

AVPlayerItem *playerItem = [AVPlayerItem playerItemWithAsset:self.composition];
playerItem.videoComposition = self.videoComposition;
self.player = [AVPlayer playerWithPlayerItem:playerItem];

AVPlayerLayer *newPlayerLayer = [AVPlayerLayer playerLayerWithPlayer:[self player]];
[newPlayerLayer setFrame:[[[self playerView] layer] bounds]];
// [newPlayerLayer setHidden:YES];

[[[self playerView] layer] addSublayer:newPlayerLayer];
[self setPlayerLayer:newPlayerLayer];

使用上面的代码我可以实现这个:

https://drive.google.com/open?id=0B2jCvCt5fosyOVNOcGZ1MU1laEU

我知道customVideoCompositor过滤组合框架的类.我尝试了但是如果我使用customVideoCompositor那么我在构图层上失去了我的转换.另外,customVideoCompositor我不知道如何过滤特定的音轨ID.

如果有人有任何文档链接或建议,那么非常感谢你继续前进.

Answer 1

实现这一目标的一种方法是使用 2 个不同的AVPlayers 和一个模糊覆盖视图：backgroundLayer-> bluer overlay view-> frontLayer。您只需确保两个播放器同时开始和停止即可。

另一种选择是使用 1AVPlayer和 a time observer。提取每一帧的当前图像frontLayer，对其进行模糊处理并以backgroundLayer. 模糊功能可以在我上面提供的同一链接中找到。