Plucko AI Grape-Picking Robot
Description: The Problem & Solution
Grape harvesting is a labor-intensive and delicate process. "Plucko" is designed to fully automate this task using AI and robotics. I contributed to all aspects of the project, from mechanical design to developing the robot's "eyes"—an advanced computer vision system for grape detection.
Key Feature: The AI Vision System
To power the robot, a precise vision system was essential. Based on work documented in the Grape Detection Project, the process included:
github.com/mostafamhialden/Graduation-project
- Data Collection: Gathering an initial dataset of grape images in various environments.
- Data Annotation: Manually annotating thousands of images to accurately label grape clusters.
- Data Augmentation: Applying techniques (cropping, rotation, lighting changes) to increase data volume and ensure model robustness.
Features & Technologies
Mechanical Design
Designing the robot's chassis and arm, applying Forward/Inverse Kinematics (FK/IK).
Deep Learning Detection
Training and experimenting with multiple deep learning models to achieve high-accuracy detection in varied conditions.
Control Systems
Developing intelligent control systems linking "vision" (input) to the robotic arm's "action" (output).
Full Integration
Merging the AI vision system with the physical robotics platform into one seamless, functional system.
Technologies Used
- Python (Core Development)
- OpenCV (Image Processing)
- TensorFlow / Keras (Deep Learning)
- Robotics (Control Systems, Kinematics)
- 3D Printing (Prototyping)
- Data Annotation & Augmentation