How to Upgrade

From version 1.x to 2.x

DynamsoftDocumentNormalizer SDK has been revamped to integrate with DynamsoftCaptureVision (DCV) architecture. To upgrade from version 1.x to 2.x, we recommend that you follow the User Guide and re-write your code to match what is in the guide.

Here are some of the main actions you may take:

Update the Dependencies

Since the SDK architecture has changed, you have to change your Podfile to include the following libraries:

pod 'DynamsoftDocumentNormalizerBundle','{version-number}'

Note: Please view user guide for the most up-to-date version number.

Migrate from Class DynamsoftDocumentNormalizer to Class DSCaptureVisionRouter

The DSCaptureVisionRouter class serves as the central class of DCV’s execution flow. It encompasses the following functionalities:

• Retrieving images from the DSImageSourceAdapter.
• Updating templates and configuring settings.
• Dynamically loading the DynamsoftDocumentNormalizer module for document detection and normalization.
• Dispatching the results to registered receivers of type DSCapturedResultReceiver.

Convert Parameter Templates

The parameter system has been restructured and the template you used for v1.x cannot be directly recognized by v2.x (and beyond). Please contact us to upgrade your parameter template.

Update the Code

Single Image Processing

You should now utilize the provided multiple capture API methods instead of the previous detectQuad and normalize API methods. These capture API methods directly return results for boundary detection or document normalization.

• Objective-C
• Swift

1. - (nullable DSCapturedResult *)captureFromFile:(NSString *)file
                             templateName:(nonnull NSString*)templateName
                                    error:(NSError *_Nullable *_Nullable)error;
   - (nullable DSCapturedResult *)captureFromFileBytes:(NSData *)fileBytes
                                             templateName:(nonnull NSString*)templateName
                                                    error:(NSError *_Nullable *_Nullable)error;
   - (nullable DSCapturedResult *)captureFromBuffer:(DSImageData *)buffer
                               templateName:(nonnull NSString*)templateName
                                      error:(NSError * _Nullable * _Nullable)error;
   - (nullable DSCapturedResult *)captureFromImage:(UIImage *)image
                               templateName:(nonnull NSString*)templateName
                                       error:(NSError *_Nullable *_Nullable)error;
   

2. func captureFromFile(_ file:String, templateName:String) throws -> CapturedResult
   func captureFromFileBytes(_ fileBytes:Data, templateName:String) throws -> CapturedResult
   func captureFromBuffer(_ buffer:DSImageData, templateName:String) throws -> CapturedResult
   func captureFromImage(_ image:UIImage, templateName:String) throws -> CapturedResult
   

Batch Image Processing

The DCV architecture allows you to conveniently and continuously obtain frames from the DSCameraEnhancer, and then detect or normalize them. The key steps are as follows:

• Set a DSCameraEnhancer object as the input of the DSCaptureVisionRouter object via the setInput API method.
• Register a DSCapturedResultReceiver object as the output of the DSCaptureVisionRouter object via the addResultReceiver API method.
• Open/Close camera via DSCameraEnhancer.open and DSCameraEnhancer.close API methods.
• Start/Stop capturing via DSCaptureVisionRouter.startCapturing and DSCaptureVisionRouter.stopCapturing API methods.

• Objective-C
• Swift

1. @interface ViewController ()<DSCapturedResultReceiver>
   @property (nonatomic, strong) DSCaptureVisionRouter *cvr;
   @property (nonatomic, strong) DSCameraEnhancer *dce;
   @property (nonatomic, strong) DSCameraView *cameraView;
   @end
   @implementation ViewController
   - (void)viewDidLoad {
   NSLog(@"View Did load");
   [super viewDidLoad];
   [self setUpCamera];
   [self setUpCvr];
   _implementCapture = false;
   [self setUpUI];
   }
   - (void)viewWillAppear:(BOOL)animated
   {
   [super viewWillAppear:animated];
   [_dce open];
   NSError *cvrError;
   [_cvr startCapturing:DSPresetTemplateDetectAndNormalizeDocument error:&cvrError];
   }
   - (void)viewWillDisappear:(BOOL)animated
   {
   [super viewWillAppear:animated];
   [_dce close];
   }
   - (void)setUpCamera
   {
   _cameraView = [[DSCameraView alloc] initWithFrame:self.view.bounds];
   [_cameraView setAutoresizingMask:UIViewAutoresizingFlexibleWidth];
   [self.view addSubview:_cameraView];
   [_cameraView addSubview:_captureButton];
   _dce = [[DSCameraEnhancer alloc] init];
   [_dce setCameraView:_cameraView];
   DSDrawingLayer * layer = [_cameraView getDrawingLayer:DSDrawingLayerIdDDN];
   [layer setVisible:true];
   }
   - (void)setUpCvr
   {
   _cvr = [[DSCaptureVisionRouter alloc] init];
   _allCornersInOneScan = [[NSMutableArray alloc] init];
   NSError *cvrError;
   [_cvr setInput:_dce error:&cvrError];
   [_cvr addResultReceiver:self error:&cvrError];
   DSMultiFrameResultCrossFilter *filter = [[DSMultiFrameResultCrossFilter alloc] init];
   [filter enableResultCrossVerification:DSCapturedResultItemTypeNormalizedImage isEnabled:true];
   [_cvr addResultFilter:filter error:&cvrError];
   }
   - (void)setUpUI {
   [self.view addSubview:self.captureButton];
   }
   -(void)onNormalizedImagesReceived:(DSNormalizedImagesResult *)result
   {
      // Add your code to do when NormalizedImagesResult is receive.
   }
   @end
   

2. class ViewController: UIViewController, CapturedResultReceiver {
   var cameraView:CameraView!
   var dce:CameraEnhancer!
   var cvr:CaptureVisionRouter!
   override func viewDidLoad() {
       super.viewDidLoad()
       setUpCamera()
       setUpDCV()
       addCaptureButton()
   }
   override func viewWillAppear(_ animated: Bool) {
       super.viewWillAppear(animated)
       dce.open()
       try? cvr.startCapturing(PresetTemplate.detectAndNormalizeDocument.rawValue)
   }
   override func viewWillDisappear(_ animated: Bool) {
       super.viewWillDisappear(animated)
       dce.close()
   }
   func setUpCamera() {
       cameraView = .init(frame: view.bounds)
       cameraView.autoresizingMask = [.flexibleWidth, .flexibleHeight]
       view.insertSubview(cameraView, at: 0)
       // Bind the camera enhancer with the camera view.
       dce = CameraEnhancer()
       dce.cameraView = cameraView
       // Additional step: Highlight the detected document boundary.
       let layer = cameraView.getDrawingLayer(DrawingLayerId.DDN.rawValue)
       layer?.visible = true
   }
   func setUpDCV() {
       cvr = CaptureVisionRouter()
       try? cvr.setInput(dce)
       try? cvr.addResultReceiver(self)
   }
   func onNormalizedImagesReceived(_ result: NormalizedImagesResult) {
       // Add your code to do when NormalizedImagesResult is receive.
   }
   }