Cirkit Designer Logo
Cirkit Designer
Your all-in-one circuit design IDE
Home / 
Component Documentation

How to Use huskylense: Examples, Pinouts, and Specs

Image of huskylense
Cirkit Designer LogoDesign with huskylense in Cirkit Designer

Introduction

HuskyLens is an AI-powered camera module designed for object recognition, face detection, color recognition, and more. It features a user-friendly interface with a built-in display, making it easy to configure and visualize results in real-time. HuskyLens is powered by advanced machine learning algorithms, enabling it to perform tasks such as object tracking, line following, and tag recognition without requiring external processing.

This versatile module is widely used in robotics, IoT projects, and educational applications. Its compatibility with microcontrollers like Arduino, Raspberry Pi, and other platforms makes it an excellent choice for developers and hobbyists alike.

Explore Projects Built with huskylense

Use Cirkit Designer to design, explore, and prototype these projects online. Some projects support real-time simulation. Click "Open Project" to start designing instantly!
Arduino Uno R3 Controlled Pan-Tilt Security Camera with Night Vision
Image of MOTION CAMERA: A project utilizing huskylense in a practical application
This circuit features an Arduino Uno R3 microcontroller connected to a Huskylens (an AI camera module), an IR LED Night Vision Ring, and a Tilt Pan module. The Huskylens is interfaced with the Arduino via I2C communication using the SDA and SCL lines, while the Tilt Pan module is controlled by the Arduino through digital pins 10 and 11 for signal and output control. The IR LED ring and Tilt Pan are powered directly from the Arduino's 5V output, and all components share a common ground.
Cirkit Designer LogoOpen Project in Cirkit Designer
Arduino UNO Controlled RGB LED and Servo System with Capacitive Sensing and HuskyLens Vision
Image of orca: A project utilizing huskylense in a practical application
This circuit features an Arduino UNO microcontroller connected to a Huskylens for image processing, a capacitive sensor for touch input, and a multi-channel PWM servo shield controlling several servos. The Arduino is powered by 5V and shares a common ground with the Huskylens and capacitive sensor, which also interface with the Arduino's analog and I2C pins, respectively. The servos are powered by a battery through a DC-DC step-down converter, and their control signals are managed by the PWM servo shield, which is also connected to the Arduino for I2C communication.
Cirkit Designer LogoOpen Project in Cirkit Designer
Raspberry Pi 5 Motion-Activated Dual DC Motor System with PIR Sensors
Image of Bhuvan: A project utilizing huskylense in a practical application
This circuit uses a Raspberry Pi 5 to control two DC motors via an L298N motor driver, based on input from two PIR motion sensors. The Raspberry Pi also interfaces with a Huskylens for additional sensor input and a 7-inch WaveShare display for output. Power is supplied by a 12V battery for the motor driver and a 5V battery for the display.
Cirkit Designer LogoOpen Project in Cirkit Designer
ESP8266 Smart Dustbin with Ultrasonic and IR Sensors
Image of Smart Dustbin: A project utilizing huskylense in a practical application
This circuit is a smart dustbin system that uses an ESP8266 microcontroller to control an ultrasonic sensor for measuring the dustbin level, an IR sensor for obstacle detection, and a servo motor to open and close the dustbin lid. The system is powered via a USB power source and operates by opening the lid when an obstacle is detected and measuring the distance to determine the fill level of the dustbin.
Cirkit Designer LogoOpen Project in Cirkit Designer

Explore Projects Built with huskylense

Use Cirkit Designer to design, explore, and prototype these projects online. Some projects support real-time simulation. Click "Open Project" to start designing instantly!
Image of MOTION CAMERA: A project utilizing huskylense in a practical application
Arduino Uno R3 Controlled Pan-Tilt Security Camera with Night Vision
This circuit features an Arduino Uno R3 microcontroller connected to a Huskylens (an AI camera module), an IR LED Night Vision Ring, and a Tilt Pan module. The Huskylens is interfaced with the Arduino via I2C communication using the SDA and SCL lines, while the Tilt Pan module is controlled by the Arduino through digital pins 10 and 11 for signal and output control. The IR LED ring and Tilt Pan are powered directly from the Arduino's 5V output, and all components share a common ground.
Cirkit Designer LogoOpen Project in Cirkit Designer
Image of orca: A project utilizing huskylense in a practical application
Arduino UNO Controlled RGB LED and Servo System with Capacitive Sensing and HuskyLens Vision
This circuit features an Arduino UNO microcontroller connected to a Huskylens for image processing, a capacitive sensor for touch input, and a multi-channel PWM servo shield controlling several servos. The Arduino is powered by 5V and shares a common ground with the Huskylens and capacitive sensor, which also interface with the Arduino's analog and I2C pins, respectively. The servos are powered by a battery through a DC-DC step-down converter, and their control signals are managed by the PWM servo shield, which is also connected to the Arduino for I2C communication.
Cirkit Designer LogoOpen Project in Cirkit Designer
Image of Bhuvan: A project utilizing huskylense in a practical application
Raspberry Pi 5 Motion-Activated Dual DC Motor System with PIR Sensors
This circuit uses a Raspberry Pi 5 to control two DC motors via an L298N motor driver, based on input from two PIR motion sensors. The Raspberry Pi also interfaces with a Huskylens for additional sensor input and a 7-inch WaveShare display for output. Power is supplied by a 12V battery for the motor driver and a 5V battery for the display.
Cirkit Designer LogoOpen Project in Cirkit Designer
Image of Smart Dustbin: A project utilizing huskylense in a practical application
ESP8266 Smart Dustbin with Ultrasonic and IR Sensors
This circuit is a smart dustbin system that uses an ESP8266 microcontroller to control an ultrasonic sensor for measuring the dustbin level, an IR sensor for obstacle detection, and a servo motor to open and close the dustbin lid. The system is powered via a USB power source and operates by opening the lid when an obstacle is detected and measuring the distance to determine the fill level of the dustbin.
Cirkit Designer LogoOpen Project in Cirkit Designer

Technical Specifications

  • Processor: Kendryte K210 AI chip
  • Display: 2.0-inch IPS screen
  • Camera Resolution: 2 MP
  • Communication Interfaces: UART, I2C
  • Input Voltage: 3.3V to 5V
  • Power Consumption: ~200mA at 5V
  • Dimensions: 52mm x 44mm x 20mm
  • Weight: ~30g
  • Supported Algorithms:
    • Face recognition
    • Object recognition
    • Object tracking
    • Line tracking
    • Color recognition
    • Tag (QR code) recognition

Pin Configuration and Descriptions

The HuskyLens module has a 4-pin interface for communication and power. The pinout is as follows:

Pin Name Description
1 VCC Power input (3.3V to 5V)
2 GND Ground
3 TX (UART) UART Transmit (data output)
4 RX (UART) UART Receive (data input)

For I2C communication, the TX and RX pins are repurposed as follows:

Pin Name Description
3 SCL I2C Clock Line
4 SDA I2C Data Line

Usage Instructions

Connecting HuskyLens to an Arduino UNO

To use HuskyLens with an Arduino UNO, follow these steps:

  1. Wiring:

    • Connect the VCC pin of HuskyLens to the 5V pin on the Arduino.
    • Connect the GND pin of HuskyLens to the GND pin on the Arduino.
    • For UART communication:
      • Connect the TX pin of HuskyLens to the RX pin (pin 0) on the Arduino.
      • Connect the RX pin of HuskyLens to the TX pin (pin 1) on the Arduino.
    • For I2C communication:
      • Connect the SCL pin of HuskyLens to the A5 pin on the Arduino.
      • Connect the SDA pin of HuskyLens to the A4 pin on the Arduino.
  2. Install Libraries:

    • Download and install the HuskyLens Arduino library from the official DFRobot GitHub repository or the Arduino Library Manager.
  3. Upload Example Code: Use the following example code to test the connection and retrieve data from HuskyLens:

    #include "HUSKYLENS.h" // Include the HuskyLens library
    
    HUSKYLENS huskylens; // Create a HuskyLens object
    
    void setup() {
      Serial.begin(9600); // Initialize serial communication
      while (!Serial);    // Wait for the serial monitor to open
    
      // Initialize HuskyLens with I2C communication
      if (!huskylens.begin(Wire)) {
        Serial.println("HuskyLens initialization failed!");
        while (1);
      }
      Serial.println("HuskyLens initialized successfully!");
    }
    
    void loop() {
      if (huskylens.request()) { // Request data from HuskyLens
        HUSKYLENSResult result = huskylens.read(); // Read the result
        if (result.command == COMMAND_RETURN_BLOCK) {
          // Print object ID and coordinates
          Serial.print("Object ID: ");
          Serial.print(result.ID);
          Serial.print(", X: ");
          Serial.print(result.xCenter);
          Serial.print(", Y: ");
          Serial.println(result.yCenter);
        }
      } else {
        Serial.println("No data received from HuskyLens.");
      }
      delay(100); // Add a short delay between requests
    }
    

Important Considerations

  • Ensure the HuskyLens module is powered within the specified voltage range (3.3V to 5V).
  • Use level shifters if connecting to a 3.3V microcontroller to avoid damaging the module.
  • When using UART communication, avoid conflicts with the Arduino's serial monitor, as both use the same pins.
  • For optimal performance, ensure the camera lens is clean and unobstructed.

Troubleshooting and FAQs

Common Issues

  1. HuskyLens not initializing:

    • Ensure the wiring is correct and the module is receiving power.
    • Verify that the correct communication protocol (UART or I2C) is selected in the code.
  2. No data received from HuskyLens:

    • Check the baud rate for UART communication (default is 9600).
    • Ensure the HuskyLens is in the correct mode (e.g., object recognition mode).
  3. Unstable or incorrect recognition results:

    • Ensure proper lighting conditions for the camera.
    • Re-train the HuskyLens for better accuracy.

FAQs

  1. Can HuskyLens work with Raspberry Pi?

    • Yes, HuskyLens supports UART and I2C communication, making it compatible with Raspberry Pi.
  2. How many objects can HuskyLens recognize simultaneously?

    • HuskyLens can recognize multiple objects simultaneously, depending on the selected algorithm.
  3. Can I update the firmware on HuskyLens?

    • Yes, firmware updates can be performed using the official HuskyLens software and a USB connection.
  4. What is the maximum detection range of HuskyLens?

    • The detection range depends on the object size and lighting conditions but typically ranges from 0.5m to 2m.

By following this documentation, you can effectively integrate HuskyLens into your projects and troubleshoot common issues.