Image Processing

Available on: iOS Android

Overview

The SDK provides powerful image processing capabilities that transform raw camera captures into high-quality, readable documents. These capabilities include document detection, perspective correction, image enhancement, and quality analysis.

Document Detection

Document edge detection analyzes an image and returns a quadrangle representing the four corners of the detected document.

                    import GSSDK

let documentDetector = GSKDocumentDetector(configuration: .defaultConfiguration)
let result = try documentDetector.detectDocumentInImage(image, options: [])
let quadrangle = result.quadrangle

                    import com.geniusscansdk.core.DocumentDetector

val detector = DocumentDetector.create(context)

// From image file
val quadrangle = detector.detectDocument(imageFile)

// Or from image bitmap
val bitmap = BitmapFactory.decodeFile(imagePath)
val quadrangle = detector.detectDocument(bitmap)
…

                  

Document Processing

The document processor takes an image and applies various enhancement operations. The SDK supports multiple processing operations that can be combined to achieve optimal results.

Processing Operations

Operation	Description	Use Case
Perspective Correction	Corrects the viewing angle	Documents photographed at an angle
Curvature Correction	Flattens curved documents	Books, magazines, curved papers
Enhancement Filters	Improve readability	Various document types
Rotation	Auto-rotate to proper orientation	Mixed orientation captures
Readability Detection	Detect blur and quality issues	Quality assurance

Implementation

                    import GSSDK

let image = // Your captured image
let quadrangle = // Detected document boundaries

// Configure processing operations
let perspectiveCorrectionConfiguration = GSKPerspectiveCorrectionConfiguration.withQuadrangle(quadrangle)
let curvatureCorrectionConfiguration = GSKCurvatureCorrectionConfiguration.withCurvatureCorrection(true)
let enhancementConfiguration = GSKEnhancementConfiguration.withFilter(.blackAndWhite)

// Process the image
let result = try GSKScanProcessor().processImage(
    image,
    perspectiveCorrectionConfiguration: perspectiveCorrectionConfiguration,
    curvatureCorrectionConfiguration: curvatureCorrectionConfiguration,
    enhancementConfiguration: enhancementConfiguration,
    rotationConfiguration: .automatic(),
    readabilityConfiguration: .enabled(),
    outputConfiguration: .defaultConfiguration
)

// Access results
let processedImage = UIImage(contentsOfFile: result.processedImagePath)

                  

                    import com.geniusscansdk.core.*

val processor = ScanProcessor(context)

// Configure processing
val configuration = ScanProcessor.Configuration(
    perspectiveCorrection = PerspectiveCorrection.automatic(),
    enhancement = Enhancement.automatic(),
    curvatureCorrection = true,
    rotation = Rotation.automatic(),
    outputConfiguration = OutputConfiguration.file(outputFolder)
}

// Process the image
val result = processor.process(imageFile, configuration)

// Access results
val processedImageFile = result.output

                  

Enhancement Filters

The SDK provides several enhancement filters optimized for different document types:

Filter	Description	Best For
No Op	No enhancement applied	Already good quality images
Strong Monochrome	Strong document filter resulting in a monochrome document	Text documents, forms, high contrast needed
Strong Grayscale	Strong document filter resulting in a grayscale document	Documents needing strong enhancement without color
Strong Color	Strong document filter resulting in a color document	Colored documents needing strong enhancement
Soft Grayscale	Soft document filter resulting in a grayscale document	Documents with subtle details, lighter enhancement
Soft Color	Soft document filter resulting in a color document	Colored documents with lighter enhancement
Photo	Filter specially designed for photos	Photos
Dark Background	Filter for documents with dark backgrounds	White text on black background, inverted documents

What’s Next?

After processing your images, you can extract text using OCR:

Next: OCR →

Image Processing

Overview

Document Detection

Document Processing

Processing Operations

Implementation

Enhancement Filters

What’s Next?

Products

Industries

Case Studies

Integration

Company