Pick and Place System

A robotic pick and place system using XArm robot and Intel RealSense camera with computer vision object detection.

XArm robotic arm with Intel RealSense camera and colored blocks for pick and place operations

Features

• Real-time object detection using Roboflow API
• Intel RealSense camera integration
• XArm robotic arm control
• Automatic object centering and picking
• Computer vision-based rotation detection

Object Detection Example

The system uses computer vision to detect and classify objects with bounding boxes:

Example showing detected red and green boxes with bounding boxes and classification labels

Requirements

• XArm robotic arm
• Intel RealSense camera
• Python 3.7+

Installation

1. Clone this repository
2. Install dependencies:

pip install -r requirements.txt

Usage

1. Ensure your XArm robot is connected and accessible at IP 192.168.1.210
2. Connect your Intel RealSense camera
3. Run the main program:

python main.py

Configuration

Update the following settings in src/pick_and_place/pick_and_place.py:

• Robot IP address
• Roboflow API key
• Home, place, and safe positions
• Camera calibration parameters

Project Structure

pick-and-place/
├── main.py                         # Entry point
├── requirements.txt                # Dependencies
├── README.md                      # Documentation
├── LICENSE                        # License
└── src/
    └── pick_and_place/
        ├── __init__.py            # Package initialization
        ├── __about__.py           # Package metadata
        ├── camera.py              # RealSense camera interface
        └── pick_and_place.py      # Main pick and place logic

Classes

RealSenseCamera

• get_frame(): Capture image from camera
• show_image(image): Display image
• release(): Stop camera

PickAndPlace

• run(): Main execution loop
• detect_objects(): Object detection using Roboflow
• center_object_in_frame(): Center object in camera view
• find_rotation(): Calculate object rotation angle
• convert_coordinates(): Convert image to robot coordinates