SafeGate iOS SDK

The SafeGate iOS SDK, for integrating SafeGate device into your iOS application. Including: - Face, Glasses, Masks, Distance Detection - SafeGate Thermal Camera Connection - SafeGate Thermal Camera Measurement

• Requirements
• Installation
  • CocoaPods
• Update Info.plist
  • Bluetooth Usage
• Usage
  • Connection
  • Measurement
  • Detections
  • HeatMap

• Apple Silicon Support

• License

Requirements

• iOS 12.0+
• Xcode 11.0+

Installation

CocoaPods

1. Create a Podfile if you don't have one: pod init
2. Add SafeGateSDK to your Podfile: pod 'SafeGateSDK', '~> 1.0.0'
3. Save the file and run: pod install. This creates an .xcworkspace file for your app. Use this file for all future development on your application.

Update Info.plist

Bluetooth Usage:

SDK uses Bluetooth to estabilish connection with Thermal Camera. When installing SafeGateSDK, you'll need to make sure that you have a NSBluetoothAlwaysUsageDescription and NSBluetoothPeripheralUsageDescription entries in your Info.plist.

Usage

SDK works with this main classes:

• ThermalCamera - camera instance class with id, name, isConnected state and peripheral - CoreBluetooth peripheral class.
• TemperatureMeter - helps to connect, to listen connection state of ThermalCamera, measure and work with Temperatures received from camera.
• Temperature/Temperatures - type aliases of Measurement<UnitTemperature> and [[Temperature]]
• HeatMapDrawer - draws HeatMap image from temperatures data with setted Pallete.
• Detection - helps to detect Face from VideoImageBuffer, to detect Mask, Glasses at face, to detect Distance to face.

In your file:

import SafeGateSDK

Connection

Create TemperatureMeter instance by default init:

 let temperatureMeter = TemperatureMeter()

Available 3 connection options for thermal camera:

1. Connect with thermal camera by id:

func estabilishConnectionWithCamera(id: UUID, timeout: Int, completion: @escaping Completion<ThermalCamera>)

 temperatureMeter.estabilishConnectionWithCamera(id: cameraId, timeout: 10) { result in
    switch result {
    case .success(let camera):
       print(camera.id)
    case .failure(ler error):
       print(error)
    }
}

2. Connect with first found camera:

func estabilishConnectionWithFirstThermalCamera(timeout: Int, completion: @escaping Completion<ThermalCamera>)

 temperatureMeter.estabilishConnectionWithFirstThermalCamera(timeout: 10) { result in
    switch result {
    case .success(let camera):
       print(camera.id)
    case .failure(ler error):
       print(error)
    }
}

3. Restore connection with already saved camera:

func connect(to camera: ThermalCamera, completion: @escaping Completion<ThermalCamera>)

 temperatureMeter.connect(to: savedCamera) { result in
    switch result {
    case .success(let camera):
       print(camera.id)
    case .failure(ler error):
       print(error)
    }
}

Completion of each func returns result with connected ThermalCamera or Error if connection failed.

Measurement

Create TemperatureMeter instance by default init:

 let temperatureMeter = TemperatureMeter()

Before measurement you must connect to thermal camera. All functions with measurement at SDK returns in deciKelvin(Kelvin * 10^-1) Measure temperature from the camera:

func measure(completion: @escaping Completion<Temperatures>)

 temperatureMeter.measure { result in
    switch result {
    case .success(let temperatures):
       print(temperatures)
    case .failure(ler error):
       print(error)
    }
}

Completion returns Temperatures or Error if measurement failed. Temperatures is 2d array, it's representing 32x32 pixels image with temperatures. Each item of array or pixel of image is deciKelvin temperature.

HeatMap

HeatMap drawing documentation will be added later.

Detections

Create Detection instance by default init:

 let detectionManager = Detection()

Detect Face from Video Image Buffer:

func detectFace(at buffer: CMSampleBuffer, metadata: Metadata, completion: @escaping Completion<Face?>)

 detectionManager.detectFace(at: imageBuffer, metadata: metadata) { result in
    switch result {
    case .success(let face):
	    if let face = face {
            print(face)
	    } else {
		    print("not found face")
	    }
    case .failure(ler error):
        print(error)
    }
}

Detect Mask, Glasses on face, Distance to face:

func detectGlasses(at face: Face, classifier: Classifier.Type?, confidence: Float, completion: @escaping Completion<Glasses>)

func detectMask(at face: Face, classifier: Classifier.Type?, confidence: Float, completion: @escaping Completion<Mask>)

func detectDistance(to face: Detection.Face, normalRange range: ClosedRange<Distance.Value>, completion: @escaping Completion<Distance>)

Apple Silicon Support

The SafeGate iOS SDK does not support Apple Silicon arm64 architecture. For Cocoapods integrations, we have explicitly excluded arm64 architecture from simulator builds to prevent build failures. We do this by modifying app and pod build settings via the Intercom podspec. We hope to be able to remove these build setting requirements and provide support for Apple Silicon in the near future when we distribute the iOS SDK as an XCFramework binary.