ios - ARKit 1.5 : Extract the recognized image

Question

所以，我的目标是:

• 找到已知图像
• 从 sceneView 中提取它(例如拍摄快照)
• 执行进一步处理

使用 ARReferenceImage 很容易完成第一步:

guard let referenceImages = ARReferenceImage.referenceImages(inGroupNamed: "AR Resources", bundle: nil) else { return }

let configuration = ARWorldTrackingConfiguration()

configuration.detectionImages = referenceImages

但现在我不知道如何从 SceneView 中提取图像。我将平面节点添加到 imageAnchor:

    func renderer(_ renderer: SCNSceneRenderer, didAdd node: SCNNode, for anchor: ARAnchor) {
    DispatchQueue.main.async { [unowned self] in
        guard let imageAnchor = anchor as? ARImageAnchor
            else { return }

        let planeNode = self.getPlaneNode(withReferenceImage: imageAnchor.referenceImage)
        planeNode.opacity = 0.5
        planeNode.eulerAngles.x = -.pi / 2
        node.addChildNode(planeNode)
    }
}

结果:

因此，我需要平面节点在屏幕上的投影来获取节点的二维屏幕坐标，然后切割图像。我找到了这个方法:renderer.projectPoint(node.position) 但它对我没有帮助。当整张图片都显示在屏幕上时，它甚至可以返回负值。

我的做法是否正确？任何帮助将不胜感激.

Answer 1

就我而言，我解决了它。

    let transform = node.simdConvertTransform(node.simdTransform, to: node.parent!)

    let x = transform.columns.3.x
    let y = transform.columns.3.y
    let z = transform.columns.3.z

    let position = SCNVector3(x, y, z)
    let projection = renderer.projectPoint(position)

    let screenPoint =  CGPoint(x: CGFloat(projection.x), y: CGFloat( projection.y))

更多信息

ARKit 2 - Crop Recognizing Images

iOS 8 Core Image saving a section of an Image via Swift

https://nacho4d-nacho4d.blogspot.com/2012/03/coreimage-and-uikit-coordinates.html