Dewarp passport

[blinkid]

Dewarp passport

[DoDoENT]

You can try this example, but also enable returning of dewarped image as described here.

[vladimirzivanov]

How to see the image which is scaning?

[DoDoENT]

After obtaining PPImageMetadata as explained in link above, you can get UIImage from its image property. You can then display this UIImage as you wish.

[vladimirzivanov]

Thanks. Can i somehow get whole image which is procesing (from camera, not croped) and get bounds of object if is it recognized(in coordinate)?

[DoDoENT]

Yes, you can. To obtain every video frame as PPImageMetadata, enable receiving of currentFrame. To obtain frame that was used to obtain successful recognition, enable receiving of successfulFrame.

[vladimirzivanov]

Thanks for all. And how to get coordinates of recognized object on successful frame? Is it possible?

[DoDoENT]

Yes, it is possible. You just need to implement scanningViewController:didFinishDetectionWithResult:

[zjoka]

We are trying to understand how PPDetectorResult works. For example if we use PPQuadDetectorResult.detectionLocation to get coordinate points, how to actually use return value to show corner points on screen. Example code below give us strange results.

override func cameraViewController(_ cameraViewController: UIViewController, didFinishDetectionWith result: PPDetectorResult) {
        if let docResult = result as? PPQuadDetectorResult {
            if result.status == PPDetectionStatus.success{
                let p = docResult.detectionLocation
                topLeftView.frame = CGRect(x: p.upperLeft.x, y: p.upperLeft.y, width: 5, height: 5)
                topRightView.frame = CGRect(x: p.upperRight.x, y: p.upperRight.y, width: 5, height: 5)
                bottomLeftView.frame = CGRect(x: p.lowerLeft.x, y: p.lowerLeft.y, width: 5, height: 5)
                bottomRightView.frame = CGRect(x: p.lowerRight.x, y: p.lowerRight.y, width: 5, height: 5)
            }
        }
    }

[DoDoENT]

The coordinates in PPQuadDetectorResult.detectionLocation are given in image's coordinate sytem. To transform those coordinates into screen coordinate system, use transform property.