Pixelbin

Pixelbin Swift library helps you integrate Pixelbin with your iOS Application.

Usage

Installing using Swift Package Manager

1. Open Your Xcode Project:
   • Launch Xcode and open your existing project or create a new one.
2. Open the Swift Packages Tab:
   • In the project navigator, select your project file to open the project settings.
   • Select the target you want to add the package to.
   • Click on the Swift Packages tab.
3. Add a Swift Package:
   • Click the + button at the bottom left of the Swift Packages tab.
   • In the search bar, enter the URL of the repository: https://github.com/pixelbin-io/pixelbin-swift-sdk.git
4. Specify Version Rules:
   • Choose the version rule that fits your needs. You can select from:
     • Branch: Specify a branch like main or master.
     • Exact Version: Specify an exact version number.
     • Range: Define a version range.
5. Add the Package:
   • Click Next to fetch the repository and validate the package.
   • After validation, select the package products you need and the target to which they should be added.
   • Click Finish to add the package to your project.

Installing using Cocoapods

• Add the dependency to your Podfile:

pod 'Pixelbin', '~> latest_version'

• Install the dependency:

pod install

• To add the Pixelbin Swift SDK as a Swift Package Manager (SPM) dependency to your iOS project using the provided URL, follow these steps:

Creating Image from URL or Cloud details

import PixelBin
guard let imageFromUrl = PixelBin.shared.image(url: "https://cdn.pixelbin.io/v2/dummy-cloudname/original/__playground/playground-default.jpeg") else {
    return
}
print(imageFromUrl.encoded)
// Create Image url from cloud, zone and other details
let imageFromDetails = PixelBin.shared.image(imagePath: imageFromUrl.imagePath, cloud: imageFromUrl.cloudName, transformations: imageFromUrl.transformations, version: imageFromUrl.version)
print(imageFromDetails.encoded)

---

Applying Transformations and Getting Transformations

import PixelBin
// Create Image url from cloud, zone and imagePath on cloud (Not local path)
let image = PixelBin.shared.image(imagePath: "example/logo/apple.jpg", cloud: "apple_cloud", zone: "south_asia")
print(imageFromDetails.encoded) // https://cdn.pixelbin.io/v2/apple_cloud/south_asia/original/example/logo/apple.jpg
let eraseTransformation = Transformation.erasebg() // Creating Erasebg Transformation
let resizeTransformation = Transformation.resize(height: 100, width: 100) // Creating Resize Transformation
image.addTransformation(eraseTransformation, resizeTransformation) // Applying transformation (as varargs, just keep passing all transformations)
let outputUrl = image.encoded // https://cdn.pixelbin.io/v2/apple_cloud/south_asia/erase.bg()~t.resize(h:100,w:100)/example/logo/apple.jpg

Uploading Image and getting url object

// Signed Url and Field can be generated via using Backed SDK for pixelbin or API to generate signed url for upload
// Assume imagePath: "example/logo/apple.jpg", cloud: "apple_cloud", zone: "south_asia" & generate details
let signUrl = "SIGNED_URL"
let fields = [] // META_DATA in value
let signedDetails = SignedDetails(url: signUrl, fields: fields)
// Url is local file path, other fields chunkSize: Int = 1024, concurrency: Int = 1
PixelBin.shared.upload(file: url, signedDetails: signedDetails) { result in
    switch result {
    case let .success(response):
        if let image = response.0 {
            print("Uploaded Image Url: \(image.encoded)")  // https://cdn.pixelbin.io/v2/apple_cloud/south_asia/original/example/logo/apple.jpg
            let eraseTransformation = Transformation.erasebg() // Creating Erasebg Transformation
            let resizeTransformation = Transformation.tResize(height: 100, width: 100) // Creating Resize Transformation
            image.addTransformation(eraseTransformation, resizeTransformation) // Applying transformation (as varargs, just keep passing all transformations)
            print("Transformed Image Url: \(image.encoded)")  // https://cdn.pixelbin.io/v2/apple_cloud/south_asia/erase.bg()~t.resize(h:100,w:100)/example/logo/apple.jpg
        }
    case let .failure(error):
        print("Error: \(error.localizedDescription)")
    }
}

Parameter	Type	Description
file (File)	File	File to upload to Pixelbin
signedDetails (SignedDetails)	Object	Signed details generated with the Pixelbin Backend SDK
chunkSize (Int)	Integer	Size of chunks to be uploaded in KB (default value is 1024)
concurrency (Int)	Integer	Number of chunks to be uploaded in parallel API calls

• Resolves with Image object on success.
• Rejects with error on failure.

List of Supported Transformations

DetectBackgroundType

1. dbtDetect()

Classifies the background of a product as plain, clean or busy

let t = Transformation.dbtDetect(
)

Basic

1. tResize(height, width, fit, background, position, algorithm, DPR)

Basic Transformations

Parameter	Type	Default
height	integer	0
width	integer	0
fit	enum: cover, contain, fill, inside, outside	Resize.Fit.cover
background	color	"000000"
position	enum: top, bottom, left, right, right_top, right_bottom, left_top, left_bottom, center	Resize.Position.center
algorithm	enum: nearest, cubic, mitchell, lanczos2, lanczos3	Resize.Algorithm.lanczos3
dpr	float	1

let t = Transformation.tResize(
    height: 0, 
    width: 0, 
    fit: Resize.Fit.cover, 
    background: "000000", 
    position: Resize.Position.center, 
    algorithm: Resize.Algorithm.lanczos3, 
    dpr: 1
)

2. tCompress(quality)

Basic Transformations

Parameter	Type	Default
quality	integer	80

let t = Transformation.tCompress(
    quality: 80
)

3. tExtend(top, left, bottom, right, background, Border Type, DPR)

Basic Transformations

Parameter	Type	Default
top	integer	10
left	integer	10
bottom	integer	10
right	integer	10
background	color	"000000"
bordertype	enum: constant, replicate, reflect, wrap	Extend.Bordertype.constant
dpr	float	1

let t = Transformation.tExtend(
    top: 10, 
    left: 10, 
    bottom: 10, 
    right: 10, 
    background: "000000", 
    bordertype: Extend.Bordertype.constant, 
    dpr: 1
)

4. tExtract(top, left, height, width, Bounding Box)

Basic Transformations

Parameter	Type	Default
top	integer	10
left	integer	10
height	integer	50
width	integer	20
boundingbox	bbox	nil

let t = Transformation.tExtract(
    top: 10, 
    left: 10, 
    height: 50, 
    width: 20, 
    boundingbox: nil
)

5. tTrim(threshold)

Basic Transformations

Parameter	Type	Default
threshold	integer	10

let t = Transformation.tTrim(
    threshold: 10
)

6. tRotate(angle, background)

Basic Transformations

Parameter	Type	Default
angle	integer	0
background	color	"000000"

let t = Transformation.tRotate(
    angle: 0, 
    background: "000000"
)

7. tFlip()

Basic Transformations

let t = Transformation.tFlip(
)

8. tFlop()

Basic Transformations

let t = Transformation.tFlop(
)

9. tSharpen(sigma)

Basic Transformations

Parameter	Type	Default
sigma	float	1.5

let t = Transformation.tSharpen(
    sigma: 1.5
)

10. tMedian(size)

Basic Transformations

Parameter	Type	Default
size	integer	3

let t = Transformation.tMedian(
    size: 3
)

11. tBlur(sigma, DPR)

Basic Transformations

Parameter	Type	Default
sigma	float	0.3
dpr	float	1

let t = Transformation.tBlur(
    sigma: 0.3, 
    dpr: 1
)

12. tFlatten(background)

Basic Transformations

Parameter	Type	Default
background	color	"000000"

let t = Transformation.tFlatten(
    background: "000000"
)

13. tNegate()

Basic Transformations

let t = Transformation.tNegate(
)

14. tNormalise()

Basic Transformations

let t = Transformation.tNormalise(
)

15. tLinear(a, b)

Basic Transformations

Parameter	Type	Default
a	integer	1
b	integer	0

let t = Transformation.tLinear(
    a: 1, 
    b: 0
)

16. tModulate(brightness, saturation, hue)

Basic Transformations

Parameter	Type	Default
brightness	float	1
saturation	float	1
hue	integer	90

let t = Transformation.tModulate(
    brightness: 1, 
    saturation: 1, 
    hue: 90
)

17. tGrey()

Basic Transformations

let t = Transformation.tGrey(
)

18. tTint(color)

Basic Transformations

Parameter	Type	Default
color	color	"000000"

let t = Transformation.tTint(
    color: "000000"
)

19. tToformat(format, quality)

Basic Transformations

Parameter	Type	Default
format	enum: jpeg, png, webp, tiff, avif, bmp, heif	Toformat.Format.jpeg
quality	enum: 100, 95, 90, 85, 80, 75, 70, 60, 50, 40, 30, 20, 10, best, good, eco, low	Toformat.Quality._75

let t = Transformation.tToformat(
    format: Toformat.Format.jpeg, 
    quality: Toformat.Quality._75
)

20. tDensity(density)

Basic Transformations

Parameter	Type	Default
density	integer	300

let t = Transformation.tDensity(
    density: 300
)

21. tMerge(mode, image, transformation, background, height, width, top, left, gravity, blend, tile, List of bboxes, List of Polygons)

Basic Transformations

Parameter	Type	Default
mode	enum: overlay, underlay, wrap	Merge.Mode.overlay
image	file	""
transformation	custom	""
background	color	"00000000"
height	integer	0
width	integer	0
top	integer	0
left	integer	0
gravity	enum: northwest, north, northeast, east, center, west, southwest, south, southeast, custom	Merge.Gravity.center
blend	enum: over, in, out, atop, dest, dest-over, dest-in, dest-out, dest-atop, xor, add, saturate, multiply, screen, overlay, darken, lighten, colour-dodge, color-dodge, colour-burn, color-burn, hard-light, soft-light, difference, exclusion	Merge.Blend.over
tile	boolean	false
listofbboxes	bboxList	nil
listofpolygons	polygonList	nil

let t = Transformation.tMerge(
    mode: Merge.Mode.overlay, 
    image: "", 
    transformation: "", 
    background: "00000000", 
    height: 0, 
    width: 0, 
    top: 0, 
    left: 0, 
    gravity: Merge.Gravity.center, 
    blend: Merge.Blend.over, 
    tile: false, 
    listofbboxes: nil, 
    listofpolygons: nil
)

Artifact

1. afRemove()

Artifact Removal Plugin

let t = Transformation.afRemove(
)

AWSRekognitionPlugin

1. awsrekDetectlabels(Maximum Labels, Minimum Confidence)

Detect objects and text in images

Parameter	Type	Default
maximumlabels	integer	5
minimumconfidence	integer	55

let t = Transformation.awsrekDetectlabels(
    maximumlabels: 5, 
    minimumconfidence: 55
)

2. awsrekModeration(Minimum Confidence)

Detect objects and text in images

Parameter	Type	Default
minimumconfidence	integer	55

let t = Transformation.awsrekModeration(
    minimumconfidence: 55
)

BackgroundGenerator

1. generateBg(Background prompt, focus, Negative prompt, seed)

AI Background Generator

Parameter	Type	Default
backgroundprompt	custom	"YSBmb3Jlc3QgZnVsbCBvZiBvYWsgdHJlZXMsd2l0aCBicmlnaHQgbGlnaHRzLCBzdW4gYW5kIGEgbG90IG9mIG1hZ2ljLCB1bHRyYSByZWFsaXN0aWMsIDhr"
focus	enum: Product, Background	BackgroundGenerator.Focus.product
negativeprompt	custom	""
seed	integer	123

let t = Transformation.generateBg(
    backgroundprompt: "YSBmb3Jlc3QgZnVsbCBvZiBvYWsgdHJlZXMsd2l0aCBicmlnaHQgbGlnaHRzLCBzdW4gYW5kIGEgbG90IG9mIG1hZ2ljLCB1bHRyYSByZWFsaXN0aWMsIDhr", 
    focus: BackgroundGenerator.Focus.product, 
    negativeprompt: "", 
    seed: 123
)

ImageExtender

1. bgExtend(Bounding Box, Prompt, Negative Prompt, Strength, Guidance Scale, Number of inference steps, Color Adjust, seed)

AI Image Extender

Parameter	Type	Default
boundingbox	bbox	nil
prompt	custom	""
negativeprompt	custom	""
strength	float	0.999
guidancescale	integer	8
numberofinferencesteps	integer	10
coloradjust	boolean	false
seed	integer	123

let t = Transformation.bgExtend(
    boundingbox: nil, 
    prompt: "", 
    negativeprompt: "", 
    strength: 0.999, 
    guidancescale: 8, 
    numberofinferencesteps: 10, 
    coloradjust: false, 
    seed: 123
)

VariationGenerator

1. vgGenerate(Generate variation prompt, No. of Variations, Seed, Autoscale)

AI Variation Generator

Parameter	Type	Default
generatevariationprompt	custom	""
noofvariations	integer	1
seed	integer	0
autoscale	boolean	true

let t = Transformation.vgGenerate(
    generatevariationprompt: "", 
    noofvariations: 1, 
    seed: 0, 
    autoscale: true
)

EraseBG

1. eraseBg(Industry Type, Add Shadow, Refine)

EraseBG Background Removal Module

Parameter	Type	Default
industrytype	enum: general, ecommerce, car, human, object	EraseBG.Industrytype.general
addshadow	boolean	false
refine	boolean	true

let t = Transformation.eraseBg(
    industrytype: EraseBG.Industrytype.general, 
    addshadow: false, 
    refine: true
)

GoogleVisionPlugin

1. googlevisDetectlabels(Maximum Labels)

Detect content and text in images

Parameter	Type	Default
maximumlabels	integer	5

let t = Transformation.googlevisDetectlabels(
    maximumlabels: 5
)

ImageCentering

1. imcDetect(Distance percentage)

Image Centering Module

Parameter	Type	Default
distancepercentage	integer	10

let t = Transformation.imcDetect(
    distancepercentage: 10
)

IntelligentCrop

1. icCrop(Required Width, Required Height, Padding Percentage, Maintain Original Aspect, Aspect Ratio, Gravity Towards, Preferred Direction, Object Type)

Intelligent Crop Plugin

Parameter	Type	Default
requiredwidth	integer	0
requiredheight	integer	0
paddingpercentage	integer	0
maintainoriginalaspect	boolean	false
aspectratio	string	""
gravitytowards	enum: object, foreground, face, none	IntelligentCrop.Gravitytowards.none
preferreddirection	enum: north_west, north, north_east, west, center, east, south_west, south, south_east	IntelligentCrop.Preferreddirection.center
objecttype	enum: airplane, apple, backpack, banana, baseball_bat, baseball_glove, bear, bed, bench, bicycle, bird, boat, book, bottle, bowl, broccoli, bus, cake, car, carrot, cat, cell_phone, chair, clock, couch, cow, cup, dining_table, dog, donut, elephant, fire_hydrant, fork, frisbee, giraffe, hair_drier, handbag, horse, hot_dog, keyboard, kite, knife, laptop, microwave, motorcycle, mouse, orange, oven, parking_meter, person, pizza, potted_plant, refrigerator, remote, sandwich, scissors, sheep, sink, skateboard, skis, snowboard, spoon, sports_ball, stop_sign, suitcase, surfboard, teddy_bear, tennis_racket, tie, toaster, toilet, toothbrush, traffic_light, train, truck, tv, umbrella, vase, wine_glass, zebra	IntelligentCrop.Objecttype.person

let t = Transformation.icCrop(
    requiredwidth: 0, 
    requiredheight: 0, 
    paddingpercentage: 0, 
    maintainoriginalaspect: false, 
    aspectratio: "", 
    gravitytowards: IntelligentCrop.Gravitytowards.none, 
    preferreddirection: IntelligentCrop.Preferreddirection.center, 
    objecttype: IntelligentCrop.Objecttype.person
)

IntelligentMasking

1. imMask(Replacement Image, Detector, Mask Type)

Intelligent Masking

Parameter	Type	Default
replacementimage	file	""
detector	enum: face, text, number_plate	IntelligentMasking.Detector.number_plate
masktype	enum: fill_black, pixelate, blur	IntelligentMasking.Masktype.fill_black

let t = Transformation.imMask(
    replacementimage: "", 
    detector: IntelligentMasking.Detector.number_plate, 
    masktype: IntelligentMasking.Masktype.fill_black
)

ObjectCounter

1. ocDetect()

Classifies whether objects in the image are single or multiple

let t = Transformation.ocDetect(
)

NSFWDetection

1. nsfwDetect(Minimum Confidence)

Detect NSFW content in images

Parameter	Type	Default
minimumconfidence	float	0.5

let t = Transformation.nsfwDetect(
    minimumconfidence: 0.5
)

NumberPlateDetection

1. numplateDetect()

Number Plate Detection Plugin

let t = Transformation.numplateDetect(
)

ObjectDetection

1. odDetect()

Detect bounding boxes of objects in the image

let t = Transformation.odDetect(
)

CheckObjectSize

1. cosDetect(Object Threshold Percent)

Calculates the percentage of the main object area relative to image dimensions.

Parameter	Type	Default
objectthresholdpercent	integer	50

let t = Transformation.cosDetect(
    objectthresholdpercent: 50
)

TextDetectionandRecognition

1. ocrExtract(Detect Only)

OCR Module

Parameter	Type	Default
detectonly	boolean	false

let t = Transformation.ocrExtract(
    detectonly: false
)

PdfWatermarkRemoval

1. pwrRemove()

PDF Watermark Removal Plugin

let t = Transformation.pwrRemove(
)

ProductTagging

1. prTag()

AI Product Tagging

let t = Transformation.prTag(
)

CheckProductVisibility

1. cpvDetect()

Classifies whether the product in the image is completely visible or not

let t = Transformation.cpvDetect(
)

QRCode

1. qrGenerate(width, height, image, margin, qRTypeNumber, qrErrorCorrectionLevel, imageSize, imageMargin, dotsColor, dotsType, dotsBgColor, cornerSquareColor, cornerSquareType, cornerDotsColor, cornerDotsType)

QRCode Plugin

Parameter	Type	Default
width	integer	300
height	integer	300
image	custom	""
margin	integer	0
qrtypenumber	integer	0
qrerrorcorrectionlevel	enum: L, M, Q, H	Generate.Qrerrorcorrectionlevel.q
imagesize	float	0.4
imagemargin	integer	0
dotscolor	color	"000000"
dotstype	enum: rounded, dots, classy, classy-rounded, square, extra-rounded	Generate.Dotstype.square
dotsbgcolor	color	"ffffff"
cornersquarecolor	color	"000000"
cornersquaretype	enum: dot, square, extra-rounded	Generate.Cornersquaretype.square
cornerdotscolor	color	"000000"
cornerdotstype	enum: dot, square	Generate.Cornerdotstype.dot

let t = Transformation.qrGenerate(
    width: 300, 
    height: 300, 
    image: "", 
    margin: 0, 
    qrtypenumber: 0, 
    qrerrorcorrectionlevel: Generate.Qrerrorcorrectionlevel.q, 
    imagesize: 0.4, 
    imagemargin: 0, 
    dotscolor: "000000", 
    dotstype: Generate.Dotstype.square, 
    dotsbgcolor: "ffffff", 
    cornersquarecolor: "000000", 
    cornersquaretype: Generate.Cornersquaretype.square, 
    cornerdotscolor: "000000", 
    cornerdotstype: Generate.Cornerdotstype.dot
)

2. qrScan()

QRCode Plugin

let t = Transformation.qrScan(
)

RemoveBG

1. removeBg()

Remove background from any image

let t = Transformation.removeBg(
)

SoftShadowGenerator

1. shadowGen(Background Image, Background Color, Shadow Angle, Shadow Intensity)

AI Soft Shadow Generator

Parameter	Type	Default
backgroundimage	file	nil
backgroundcolor	color	"ffffff"
shadowangle	float	120
shadowintensity	float	0.5

let t = Transformation.shadowGen(
    backgroundimage: nil, 
    backgroundcolor: "ffffff", 
    shadowangle: 120, 
    shadowintensity: 0.5
)

SuperResolution

1. srUpscale(Type, Enhance Face, Model, Enhance Quality)

Super Resolution Module

Parameter	Type	Default
type0	enum: 2x, 4x, 8x	SuperResolution.Type0._2x
enhanceface	boolean	false
model	enum: Picasso, Flash	SuperResolution.Model.picasso
enhancequality	boolean	false

let t = Transformation.srUpscale(
    type0: SuperResolution.Type0._2x, 
    enhanceface: false, 
    model: SuperResolution.Model.picasso, 
    enhancequality: false
)

VertexAI

1. vertexaiGeneratebg(Background prompt, Negative prompt, seed, Guidance Scale)

Vertex AI based transformations

Parameter	Type	Default
backgroundprompt	custom	"YSBmb3Jlc3QgZnVsbCBvZiBvYWsgdHJlZXMsd2l0aCBicmlnaHQgbGlnaHRzLCBzdW4gYW5kIGEgbG90IG9mIG1hZ2ljLCB1bHRyYSByZWFsaXN0aWMsIDhr"
negativeprompt	custom	""
seed	integer	22
guidancescale	integer	60

let t = Transformation.vertexaiGeneratebg(
    backgroundprompt: "YSBmb3Jlc3QgZnVsbCBvZiBvYWsgdHJlZXMsd2l0aCBicmlnaHQgbGlnaHRzLCBzdW4gYW5kIGEgbG90IG9mIG1hZ2ljLCB1bHRyYSByZWFsaXN0aWMsIDhr", 
    negativeprompt: "", 
    seed: 22, 
    guidancescale: 60
)

2. vertexaiRemovebg()

Vertex AI based transformations

let t = Transformation.vertexaiRemovebg(
)

3. vertexaiUpscale(Type)

Vertex AI based transformations

Parameter	Type	Default
type0	enum: x2, x4	Upscale.Type0.x2

let t = Transformation.vertexaiUpscale(
    type0: Upscale.Type0.x2
)

VideoWatermarkRemoval

1. wmvRemove()

Video Watermark Removal Plugin

let t = Transformation.wmvRemove(
)

VideoUpscalerPlugin

1. vsrUpscale()

Video Upscaler Plugin

let t = Transformation.vsrUpscale(
)

ViewDetection

1. vdDetect()

Classifies wear type and view type of products in the image

let t = Transformation.vdDetect(
)

WatermarkRemoval

1. wmRemove(Remove Text, Remove Logo, Box 1, Box 2, Box 3, Box 4, Box 5)

Watermark Removal Plugin

Parameter	Type	Default
removetext	boolean	false
removelogo	boolean	false
box1	string	"0_0_100_100"
box2	string	"0_0_0_0"
box3	string	"0_0_0_0"
box4	string	"0_0_0_0"
box5	string	"0_0_0_0"

let t = Transformation.wmRemove(
    removetext: false, 
    removelogo: false, 
    box1: "0_0_100_100", 
    box2: "0_0_0_0", 
    box3: "0_0_0_0", 
    box4: "0_0_0_0", 
    box5: "0_0_0_0"
)

WatermarkDetection

1. wmcDetect(Detect Text)

Watermark Detection Plugin

Parameter	Type	Default
detecttext	boolean	false

let t = Transformation.wmcDetect(
    detecttext: false
)