QuickCapture iOS SDK: Document Scanning for Native iOS Apps

QuickCapture for iOS (v4.0) is Extrieve's document scanning and imaging SDK for native iOS apps written in Swift or Objective-C. It captures documents with the device camera, then crops, optimises and compresses the image and outputs JPEG, PDF or TIFF.

Product page · Documentation · All platforms · Android demo app · Discussions · Contact

Highlights

• Built for documents: capture tuned for document quality, strong compression and image optimisation, with the readability of the output kept in mind.
• Output control: set the DPI, page layout and size of output images, and convert them to PDF and TIFF.
• QR code and barcode: scanning and generation.
• Face detection and matching: detect faces in documents and compare them on the device.
• Works offline: runs locally on the device, with no data transferred to any server or third party.
• Developer friendly: delivered as a framework for Swift and Objective-C projects.

To reduce build size, an initial internet connection may optionally be needed to fetch ML data or resource files, depending on the integration and the features used by the host application.

End of support notice

• QuickCapture iOS V1 was deprecated in June 2023.
• QuickCapture iOS V2 was deprecated in December 2024.
• Use V4 for new features and updates. V3 receives bug fixes only.

Contents

• Compatibility
• Run-time requirements
• Installation
• API and integration details
• 1. CameraHelper
• 2. Config
• 3. ImgHelper
• ImgException
• 4. HumanFaceHelper
• Notes on accuracy
• Other platforms and related SDKs
• Support

Compatibility

• Minimum IOS version: QuickCapture v4.0 requires a minimum iOS 12.
• Compatible with Swift and Objective-C
• Required Operating System: iphoneos

Run-time requirements

• This SDK is designed to run on officially supported Android & iOS devices only.
• Supported CPU architectures: arm64.
• Simulator and emulator environments are not supported.For testing on simulators, please contact the development support team to request a dedicated test version compatible with those environments.

Installation

You can download sdk as framework in .zip file from GitHub's releases page.

QuickCaptureSDK IOS v4.0

API and integration details

SDK has three core classes and supporting classes :

1. CameraHelper - Handle the camera related operations. Basically, an activity.

2. ImgHelper - Purpose of this class is to handle all imaging related operations.

3. HumanFaceHelper - Advanced Ai based utility class handles all functionalities such as face detection, extraction,matching & related functions.

4. Config - Holds various configurations of SDK.

5. ImgException - Handle all exceptions on Image related operations on ImgHelper.

Based on the requirement any one or all classes can be used. And need to import those from the SDK.

//Swift
import QuickCaptureFW;

//Objective-C

#import <QuickCaptureSDK/QuickCaptureFW.h>
#import <QuickCaptureSDK/QuickCaptureSDK-Swift.h>

---

1. CameraHelper

This class will be implemented as an activity. This class can be initialized as normal class.

//Swift

var cameraHelper = CameraHelper();

//Objective-C

CameraHelper *camerahelper = [[CameraHelper alloc] init];

With an activity call, triggering the SDK for capture activity can be done.Most operations in CameraHelper is view based.

//Swift

cameraHelper.StartCapture(
	sender:self,
	//pathtowritabledirectorywherecaptured
	//files are stored temporarily for processing
	workingDirectory:“”,
	callback:cameraHelperCallBack
)

func cameraHelperCallback(_ImageArray:[String])->Void {

//paths-  arrayofcapturedimagesavailablehere.

//Use returned images

}

//Objective-C

@property (nonatomic, copy, nullable) CaptureCallbackBlock captureCallback;

self.captureCallback = ^(NSArray<NSString *> * _Nonnull imageArray) {
		if (imageArray.count == 0) {
		NSLog(@"Error: no images captured");
	}
	else{
		///imageArray is the Argument where you get the Array of Strings
		///these strings are the path of captured image saved in WorkingDirectory temporarily
		///Note: Don't forget to delete image files after completion of task
		self->_CapturedImages = imageArray;
	}
};

//Starts Camera Capture Activity
[cameraHelper StartCapture:self :workingDir : self.captureCallback];

SDK is having multiple flows as follows :

• CAMERA_CAPTURE_REVIEW - Default flow. Capture with SDK Camera -> review.

• SYSTEM_CAMERA_CAPTURE_REVIEW - Capture with system default camera -> review.

• IMAGE_ATTACH_REVIEW - Attach/pass image -> review.

1. CAMERA_CAPTURE_REVIEW - Default flow of the CameraHelper. Includes Capture with SDK Camera -> Review Image.

//swift

Config.CaptureSupport.CaptureMode = CaptureModes.CAMERA_CAPTURE_REVIEW

//objective-c

CaptureSupport.CaptureMode = CaptureModesCAMERA_CAPTURE_REVIEW;

2. SYSTEM_CAMERA_CAPTURE_REVIEW - If user needs to capture an image with system default camera, this can be used. It includes Capture with system default camera -> Review.

//swift

Config.CaptureSupport.CaptureMode = CaptureModes.SYSTEM_CAMERA_CAPTURE_REVIEW

//objective-c

CaptureSupport.CaptureMode = CaptureModesSYSTEM_CAMERA_CAPTURE_REVIEW;

3. IMAGE_ATTACH_REVIEW - This option can be used if the user needs to review an image from their device's gallery. After attaching each image, the review and all dependent functionalities become available.

//swift

Config.CaptureSupport.CaptureMode = CaptureModes.IMAGE_ATTACH_REVIEW

//objective-c

CaptureSupport.CaptureMode = CaptureModesIMAGE_ATTACH_REVIEW;

2. Config

The SDK includes a supporting class called for static configuration. This class holds all configurations related to the SDK. It also contains sub-configuration collection for further organization.

This includes: :

CaptureSupport - Contains all the Capture & review related configurations. Config.CaptureSupport contains various configurations as follows:

• OutputPath - To set the output directory in which the captured images will be saved. Base app should have permissions to write to the provided path.

//Swift

Config.CaptureSupport.OutputPath = “pass output path as string”

//Objective-C

CaptureSupport.OutputPath = “pass output path as string”;

• MaxPage - To set the number of captures to do on each camera session. And this can also control whether the capture mode is single or multi i.e

if MaxPage <= 0 // not set: means unlimited. If MaxPage >= 1 // means limited.

// MaxPage <= 0  : Unlimited Capture Mode

// MaxPage = 1 : Limited Single Capture

// MaxPage > 1 : Limited Multi Capture Mode

//Swift

Config.CaptureSupport.MaxPage = 0

//Objective-C

CaptureSupport.MaxPage = 0;

• ColorMode - To Set the capture color mode- supporting color and grayscale.

RGB GRAY

//Swift

Config.CaptureSupport.ColorMode = ColorModes.RGB

//Objective-C

CaptureSupport.ColorMode = ColorModesRGB

• EnableFlash - Enable Document capture specific flash control for SDK camera.

//Swift

Config.CaptureSupport.EnableFlash  =  true;

//Objective-C

CaptureSupport.EnableFlash  =  true;

• EnableTorchOnLowLight - Automatically turns on the device’s torch/flashlight when low-light conditions are detected, ensuring clearer and more readable captures.

  //Swift
  
  Config.CaptureSupport.EnableTorchOnLowLight = true;

  //Objective-C
  
  CaptureSupport.EnableTorchOnLowLight = true;

• CaptureSound - To Enable camera capture sound.

//Swift

Config.CaptureSupport.CaptureSound  =  true;

//Objective-C

CaptureSupport.CaptureSound  =  true;

• ShowCaptureCountAndLimit - Enable/Disable Capture count as toast at bottom while capturing from SDK camera.

//Swift

Config.CaptureSupport.ShowCaptureCountAndLimit  =  true;

//Objective-C

CaptureSupport.ShowCaptureCountAndLimit  =  true;

• SDKInfo - Contains all version related information on SDK.

//Swift

Config.CaptureSupport.SDKInfo

//Objective-C

CaptureSupport.SDKInfo;

• CameraToggle - Toggle camera between front and back.

0 Disable camera toggle option. 1 Enable camera toggle option with Front camera by default. 2 Enable camera toggle option with Back camera by default.

//Swift

Config.CaptureSupport.CameraToggle  =  2;

//Objective-C

CaptureSupport.CameraToggle  =  2;

• BottomStampData - Allows adding custom text as a bottom stamp on captured documents.
  • $: Inserts a line break.
  • {DATETIME}: PlaceHolder that Prints the current date and time from the device.

//Swift

Config.CaptureSupport.BottomStampData = "{DATETIME} $ Extrieve Technologies";

//Objective-C

CaptureSupport.BottomStampData = "{DATETIME} $ Extrieve Technologies";

• Review - Contains review related configurations. Config.CaptureSupport.Review contains configurations as follows:
  • enableAutoCorrectDocAngle - Enables or disables AI-based automatic document rotation correction.

  //Swift
  
  Config.CaptureSupport.Review.enableAutoCorrectDocAngle = true;

  // Objective-C
  
  CaptureSupport.Review.enableAutoCorrectDocAngle = true;

3. ImgHelper

Following are the options/methods available from class ImgHelper :

To instanciate this class

//Swift

let ImgHelper = ImgHelper.GetInstance()

//Objective-C

ImgHelper *ImgHelper = [ImgHelper GetInstance];

• SetImageQuality - Set the Quality of the image, Document_Quality isused. If documents are used further for any automations and OCR, use Document_Quality.

Available Image Qualities :

1. Photo_Quality
2. Document_Quality
3. Compressed_Document

//Swift

ImgHelper.shared.SetImageQuality(value: ImageQuality.Document_Quality)

//OR

ImgHelperObj.SetImageQuality(value: ImageQuality.Document_Quality)

//Objective-C

[ImgHelperObj SetImageQuality:ImageQualityDocument_Quality];

• SetPageLayout - Set the Layout for the images generated/processed by the system.

Available layouts : A1, A2, A3, A4, A5, A6, A7,PHOTO & CUSTOM

A4 is the most recommended layout for document capture scenarios.

//swift
ImgHelper.shared.SetPageLayout(LayoutType.A4)

//OR

ImgHelper.SetPageLayout(LayoutType.A4)

//objective-c

[ImgHelperObj SetPageLayout:LayoutTypeA4];

• SetDPI - Set DPI (depth per inch) for the image.

Available DPI : DPI_150, DPI_200, DPI_300, DPI_500, DPI_600

150 & 200 DPI is most used. And 200 DPI recommended for OCR and other image extraction prior to capture.

//Swift
ImgHelper.shared.SetDPI(DPI._200)

//OR

ImgHelperObj.SetDPI(DPI._200)

//objective-c

[ImgHelperObj SetDpi:DPI_200];

• SetResizeMode - This method manages how the resizing of captured images are processed and output is created accordingly.

Available Resize Modes:

STRETCH_TO_EXACT_SIZE - Stretches the image to match exact width & height. PRESERVE_ASPECT_ONLY - Tries to fit within the given width & height to preserve quality at its best. FIT_WITH_ASPECT - Output image will be the exact size with white background colour if the captured image does not fits.

//swift

ImgHelper.shared.SetResizeMode(mode: ResizeMode.PRESERVE_ASPECT_ONLY)

//objective-c

[ImgHelper.shared SetResizeModeWithMode:ResizeModePRESERVE_ASPECT_ONLY];

• SetCustomLayOut - Sets the output image width and height, so that the output image will be resized accordingly. Accepted input parameters are in inches only.

//swift

ImgHelper.shared.SetCustomLayOut(CustomWidth: 0.8, CustomHeight: 1.1)

//objective-c

[ImgHelper.shared  SetCustomLayOutWithCustomWidth:0.8  CustomHeight:1.1];

• SetCustomLayOutInPixels - Sets the output image width and height, so that the output image will be resized accordingly. Accepted input parameters are in pixels only.

//swift

ImgHelper.shared.SetCustomLayOutInPixels(CustomWidth: 640, CustomHeight: 480)

//objective-c

[ImgHelper.shared  SetCustomLayOutInPixelsWithCustomWidth:640  CustomHeight:480];

• GetThumbnail - This method Will build thumbnail for the given image in custom width,height & AspectRatio.

//Swift

do{
	try UIImage thumb = ImgHelper.shared.GetThumbnail(bm: image,rHeight: 600,rWidth: 600,AspectRatio: true)
}
catch{
	print(error);
}
/*
UIImage GetThumbnail(
@NonNull  UIImage bm,
int reqHeight,
int reqWidth,
Boolean AspectRatio )throws ImgException.
*/

//Objective-C

UIImage *thumb = [ImageHelper GetThumbnail:image :600 :600 :true :&err];

/*
UIImage GetThumbnail(
@NonNull  UIImage image,
int reqHeight,
int reqWidth,
Boolean AspectRatio,
NSError err ).
*/

• CompressToJPEG - This method will Compress the provided bitmap image and will save to given path.

//Swift

let isCompressed = ImgHelper.shared.CompressToJPEG(image: image, outputURI: jpegURL.absoluteString)

/*
Boolean CompressToJPEG(UIImage image,String outputFilePath)
throws ImgException
*/

//Objective-C

BOOL isCompressed = [ImageHelper CompressToJPEG:image :outputJPEGURL.absoluteString];

• RotateBitmap - This method will rotate the image to preferred orientation.

//Swift

UIImage rotatedBm = ImgHelper.shared.RotateBitmap(image: img, orientation: ImageOrientation.left);

/*
UIimage RotateBitmap(UIimage image,enum ImageOrientation)
throws ImgException
Available Options for Orientation
*/
enum ImageOrientation : Int {
	up
	down
	left
	right
	upMirrored
	downMirrored
	leftMirrored
	rightMirrored
}

//Objective-C

UIImage *rotatedImage = [ImageHelper RotateBitmap:image :ImageOrientationUp :&err];

• BuildTiff - Build .tiff file output from the list of images shared.

//Swift

do{
	let status = try ImgHelper.shared.BuildTiff(ImageArray: fileArray, TiffFilePath: outputTiffURL.path)
	print(status)
}
catch{
	print(error)
}

*@param  "Image  File  path  collection  as  Array of String"

*@return  on  failure  =  "FAILED:::REASON"  ||  on  success  =  "SUCCESS:::TIFF  file  path".

//Objective-C

NSError *error = nil;
NSString *outputPath = [ImgHelper.shared BuildTiff:imageArray TiffFilePath:tiffFilePath error:&error];

*@param  "Image  File  path  collection  as  Array of String"
*@return  on  failure  =  "FAILED:::REASON"  ||  on  success  =  "SUCCESS:::TIFF  file  path".

• BuildPDF - Build PDF file output file from last captured set of images.

//Swift
do {
	let status = try ImgHelper.shared.BuildPdf(ImageArray: fileArray, PdfFilePath: outputPDFURL.path)
	print(status)
}
catch {
	print(error)
}

*@param  "Image  File  path  collection  as  Array of String"
*@return  on  failure  =  "FAILED:::REASON"  ||  on  success  =  "SUCCESS:::PDF  file  path".

//Objective-C

NSError *error = nil;
NSString *outputPath = [ImgHelper.shared BuildPdf:ImageArray PdfFilePath:pdfFilePath error:&error];

*@param  "Image  File  path  collection  as  Array of String"
*@return  on  failure  =  "FAILED:::REASON"  ||  on  success  =  "SUCCESS:::PDF  file  path".

Recommended Settings:

• ImageQuality: Document_Quality

• DPI: 150 or 200

• LayoutType: A4

• ResizeMode: PRESERVE_ASPECT_ONLY

• ReadImageInfo - Reads metadata from an image file and returns a JSON string containing information such as PageCount, Height, Width, and ColorType (COLOR, BW, or GREY).

  Supported image formats:

  • TIFF
  • JPEG / JPG
  • PNG
  • BMP

  //Swift
  
   Do{
      let jsonString = ImgHelper.shared.ReadImageInfo(inputUrl: imagepath)
      print(jsonString)
  }catch{
      print(error)
  }

  //Objective-C
  
  NSString *jsonString = [ImgHelper.shared ReadImageInfo:imagePath];
  NSLog(@"%@", jsonString);

  Parameter

  • inputUrl - Image file path as String.

  ReturnValue

  Failure Response (Returned as String containing JSON)

  {
     "Status": false,
     "ERROR_MESSAGE": "Reason for failure"
  }

  Success Response (Returned as String containing JSON)

  {
     "FILESIZE": "146 KB",
     "TOTAL_PAGES": 1,
     "INFORMATION": [
      {
    	"PAGENO": 1,
    	"WIDTH": 475,
    	"HEIGHT": 307,
    	"COLOR_TYPE": "GREY"
    	}
      ]
    }

  Field Description

  Field	Description
  FILESIZE	Total size of the image file in KB
  TOTAL_PAGES	Number of pages (TIFF may contain multiple pages)
  PAGENO	Page number of the image
  WIDTH	Width of the image in pixels
  HEIGHT	Height of the image in pixels
  COLOR_TYPE	Image color type: COLOR, GREY, or BW

ImgException

As a part of exceptional error handling ImgException class is available.

• Following are the possible errors and corresponding codes:

• CREATE_FILE_ERROR= -100;

• IMAGE_ROTATION_ERROR= -101;

• LOAD_TO_BUFFER_ERROR= -102;

• DELETE_FILE_ERROR= -103;

• GET_ROTATION_ERROR= -104;

• ROTATE_BITMAP_ERROR= -105;

• BITMAP_RESIZE_ERROR= -106;

• CAMERA_HELPER_ERROR= -107;

• LOG_CREATION_ERROR= -108;

• IMAGE_COMPRESSION_ERROR= -109;

4. HumanFaceHelper

QuickCapture SDK equipped with advanced face identification intelligence can accurately detect human faces within documents and match them precisely. SDK needs to be activated using a proper license with Config.License.Activate(); for the plus features to initialize.

Initialization Callback:

//swift
//InitializationCallback is the protocol to inherit
class YourClassName: NSObject, InitializationCallback {
    //This will be called when initialization process is success
    func onInitializationSuccess() {
        print("Initialization succeeded.")
    }
    //This will be called when initialization process is failed
    func onInitializationFailure(error: Error) {
        print("Initialization failed with error: \(error.localizedDescription)")
    }
}

//objective-c
@interface HumanFaceHelperInitCallback : NSObject <InitializationCallback>
@end

@implementation HumanFaceHelperInitCallback

- (void)onInitializationSuccess {
	NSLog(@"Initialization succeeded");
}

- (void)onInitializationFailureWithError:(NSError *)error {
	NSLog(@"Initialization failed: %@", error.localizedDescription);
}
@end

Instanciation :

//swift

var humanFaceObj: HumanFaceHelper? = nil
do{
	humanFaceObj = try HumanFaceHelper.init(mainBundle: Bundle.main, callback: self)
}
catch let error as  FaceHelperException{
	print("Error code: \(error.code), Error message: \(error.Description)")
	return
}
catch{
	print("Unknown Error Occured: \(error)")
	return
}

@Param - mainBundle //bundle of the caller application.
@Param - callback 	//object of the class which inherits InitializationCallback or self if inherited by the same class which is instanciating.
@Return 			//instance  of the HumanFaceHelper class

//objective-c

NSError *error = nil;
HumanFaceHelperInitCallback *initCallback = [HumanFaceHelperInitCallback  new];
NSBundle *bundle = [NSBundle mainBundle];
HumanFaceHelper *humanFaceObj = [[HumanFaceHelper alloc] initWithMainBundle:bundle callback:initCallback error:&error];

@Param - mainBundle //bundle of the caller application.
@Param - callback 	//object of the class which inherits InitializationCallback or self if inherited by the same class which is instanciating.
@Return 			//instance  of the HumanFaceHelper class

OR You can directly instantiate using Anonymous style callbacks within the statement

//swift

let humanFaceObj = try HumanFaceHelper(
    mainBundle: .main,
    callback: {
        class AnonymousCallback: NSObject, InitializationCallback {
            let onSuccess: () -> Void
            let onFailure: (Error) -> Void
            init(onSuccess: @escaping () -> Void, onFailure: @escaping (Error) -> Void) {
                self.onSuccess = onSuccess
                self.onFailure = onFailure
            }
            func onInitializationSuccess() {
                onSuccess()
            }
            func onInitializationFailure(error: Error) {
                onFailure(error)
            }
        }
        return AnonymousCallback(
            onSuccess: { print("✅ Init Success") },
            onFailure: { error in print("❌ Init Failed: \(error.localizedDescription)") }
        )
    }()
)

Following are the options/methods available from the class HumanFaceHelper :

• DetectHumanFaces - DetectHumanFaces Method from humanFaceObj will Identify human faces from provided image and return the detected details in json string format using response callbacks. It also supports both Anonymous style callback or inherited protocol style like we did while instantiating.

//swift

humanFaceObj.DetectHumanFaces(inputImage: yourUIImage, callback: {
	class AnonymousCallback: NSObject, ResponseCallback {
		let onSuccess: (String) -> Void
		let onFailure: (String) -> Void
		init(onSuccess: @escaping (String) -> Void, onFailure: @escaping (String) -> Void) {

			self.onSuccess = onSuccess
			self.onFailure = onFailure
		}
		func onCompleted(response: String) { onSuccess(response) }
		func onFailed(error: String) { onFailure(error) }
	}
	return AnonymousCallback(
		onSuccess: { response **in** print("Detection Completed: \(response)") },
		onFailure: { error **in** print("Detection Failed: \(error)") }
	)
}())

//objective-c

//Inside - Header file
typedef void (^SuccessBlock)(NSString *response);
typedef void (^FailureBlock)(NSString *error);

@interface AnonymousResponseCallback : NSObject <ResponseCallback>

- (instancetype)initWithSuccess:(SuccessBlock)success
                        failure:(FailureBlock)failure;
@end

//----------------------------------------------//
//Inside - objectivec file

@interface AnonymousResponseCallback ()
@property (nonatomic, copy) SuccessBlock success;
@property (nonatomic, copy) FailureBlock failure;
@end
@implementation AnonymousResponseCallback
- (instancetype)initWithSuccess:(SuccessBlock)success
                        failure:(FailureBlock)failure {
    self = [super init];
    if (self) {
        _success = [success copy];
        _failure = [failure copy];
    }
    return self;
}
- (void)onCompletedWithResponse:(NSString *)response {
    if (self.success) {
        self.success(response);
    }
}
- (void)onFailedWithError:(NSString *)error {
    if (self.failure) {
        self.failure(error);
    }
}
@end
//-------------------------------------------------------------------
//Macro definition
#define MAKE_RESPONSE_CALLBACK(successBlock, failureBlock) \
  [[AnonymousResponseCallback alloc] initWithSuccess:successBlock failure:failureBlock]

//Usage
[humanFaceObj DetectHumanFacesWithInputImage:inputUiimage
                                   callback:MAKE_RESPONSE_CALLBACK(
    ^(NSString *response) {
        NSLog(@"✅ Detection Completed: %@", response);
    },
    ^(NSString *error) {
        NSLog(@"❌ Detection Failed: %@", error);
    }
)];

Following is a sample of response structure :

{
	STATUS:  true/false,  //Detection status
	DESCRIPTION :  "SUCCESS",  //Success or failure
	description DOC_ID :  0,  //Identifier/index of the used document.
	FACE_DATA :[
	//Collection of identified face data
		{
			ID :  0,  //Identifier/index of face.
			LEFT :  0,
			TOP :  0,
			RIGHT :  0,
			BOTTOM :  0
			//Each location of face in document
		}
	]
}

• MatchHumanFaces - DetectHumanFaces Method from humanFaceObj With AI intelligence, analyses the provided face data and returns a response on whether the provided faces are of same human or not. Document Id and Face Id will be provided by DetectHumanFaces, and same can be used.

//swift

humanFaceObj!.MatchHumanFaces(
    firstDocumentId: id1,
    firstDocumentFaceIndex: index1,
    secondDocumentId: id2,
    secondDocumentFaceIndex: index2,
    callback: {
        class AnonymousCallback: NSObject, ResponseCallback {
            let onSuccess: (String) -> Void
            let onFailure: (String) -> Void

            init(onSuccess: @escaping (String) -> Void, onFailure: @escaping (String) -> Void) {
                self.onSuccess = onSuccess
                self.onFailure = onFailure
            }
            func onCompleted(response: String) {
                onSuccess(response)
            }
            func onFailed(error: String) {
                onFailure(error)
            }
        }

        return AnonymousCallback(
            onSuccess: { response in
                print("✅ Match Success: \(response)")
            },
            onFailure: { error in
                print("❌ Match Failed: \(error)")
            }
        )
    }()
)

//objective-c

//--------------------------------------------------------
//Optional if already updated for Detection skip to USAGE
//--------------------------------------------------------
//Inside - header
typedef void (^SuccessBlock)(NSString *response);
typedef void (^FailureBlock)(NSString *error);

@interface AnonymousResponseCallback : NSObject <ResponseCallback>

- (instancetype)initWithSuccess:(SuccessBlock)success
                        failure:(FailureBlock)failure;

@end

//Inside objective C file
@interface AnonymousResponseCallback ()
@property (nonatomic, copy) SuccessBlock success;
@property (nonatomic, copy) FailureBlock failure;
@end

@implementation AnonymousResponseCallback

- (instancetype)initWithSuccess:(SuccessBlock)success
                        failure:(FailureBlock)failure {
    self = [super init];
    if (self) {
        _success = [success copy];
        _failure = [failure copy];
    }
    return self;
}

- (void)onCompletedWithResponse:(NSString *)response {
    if (self.success) {
        self.success(response);
    }
}

- (void)onFailedWithError:(NSString *)error {
    if (self.failure) {
        self.failure(error);
    }
}

@end
//-----------------------------------------------------

//MACRO define before usage
#define MAKE_RESPONSE_CALLBACK(successBlock, failureBlock) \
  [[AnonymousResponseCallback alloc] initWithSuccess:successBlock failure:failureBlock]

//Usage
[humanFaceObj MatchHumanFacesWithFirstDocumentId:id1
                        firstDocumentFaceIndex:index1
                           secondDocumentId:id2
                     secondDocumentFaceIndex:index2
                                     callback:MAKE_RESPONSE_CALLBACK(
    ^(NSString *response) {
        NSLog(@"✅ Match Success: %@", response);
    },
    ^(NSString *error) {
        NSLog(@"❌ Match Failed: %@", error);
    }
)];

Following is a sample response structure :

{
	STATUS: true/false, //Match status

	DESCRIPTION : "SUCCESS",	//Success or failure description

	ACCURACY: 0,	//Accuracy of match

}

Face Match Accuracy Interpretation

The match level is determined based on the accuracy percentage, which reflects the similarity between two facial images. The table below provides detailed descriptions for each match level.

Match Percentage	Match Level	Description
90% - 100%	✅ Highly Reliable Match	Faces match with extremely high confidence. They are almost certainly the same person. Suitable for critical identification applications.
75% - 89%	✅ Strong Match	Faces matched successfully with a high probability of being the same person. Reliable for most identity verification use cases.
65% - 74%	⚠️ Moderate Match	Faces show good similarity, but further validation may be required. Manual verification is recommended before confirmation.
50% - 64%	⚠️ Low Confidence Match	Faces have some resemblance, but the similarity is not strong enough to confirm identity. Additional verification is needed.
0% - 49%	❌ No Match	Faces do not match. There is minimal similarity, and they are highly unlikely to be the same person.

Usage of Results

✅ Highly Reliable Match (90% - 100%)

• Best for: Secure identity verification, biometric authentication, and critical decision-making.

• Action: Automatic acceptance. No further review required.

✅ Strong Match (75% - 89%)

• Best for: General identification scenarios where strong confidence is required.

• Action: Safe for automatic approval in most applications.

⚠️ Moderate Match (65% - 74%)

• Best for: Cases where additional review is acceptable before finalizing the decision.

• Action: Manual verification recommended before confirming a match.

⚠️ Low Confidence Match (50% - 64%)

• Best for: Situations requiring strict validation before acceptance.

• Action: Use alternative verification methods. Do not rely on this score alone.

❌ No Match (0% - 49%)

• Best for: Definitive rejection of mismatches.

• Action: Automatically reject matches in this range.

Notes on accuracy

The accuracy of face detection and matching technologies depends on input image quality, including factors such as image distortion, rotation angles, lighting conditions, and color consistency. While offline solutions effectively reduce manual effort and operational costs, they do not guarantee 100% reliability in all scenarios.

This system enables on-device verification, efficiently identifying doubtful matches and flagging them for backend verification within the offline environment. By integrating backend validation, the system enhances reliability without relying on external APIs. Additionally, when a match achieves high accuracy as defined in the accuracy thresholds, the system can be considered reliable even without backend verification, making it a valuable solution for offline scenarios where external validation is limited.

For use cases demanding exceptionally high accuracy and reliability, an API-based advanced system is recommended.

Other platforms and related SDKs

• QuickCapture for Android, React Native, Flutter, Web
• QuickCapture overview: all platforms, features and FAQ
• SplicerAI: KYC and document data extraction, classification and Aadhaar masking
• ImageWizHelper: server and desktop image compression and conversion

Support

• Documentation: docs.extrieve.in
• Technical questions: GitHub Discussions
• Developer support: devsupport@extrieve.com
• Sales and evaluation: extrieve.com/contact

---

© 1996–2026 Extrieve Technologies. All rights reserved.