Developed an Automated Design Sorting System According to Outer Edges of Object.

Area of Science:

• Industrial Automation
• Computer Vision
• Manufacturing Technology

Background:

• Automated sorting systems enhance industrial production rates.
• Existing systems require robust methods for real-time product identification and sorting.
• Vision-based systems offer potential for flexible and accurate sorting solutions.

Purpose of the Study:

• To develop and evaluate an automated vision system for real-time product sorting.
• To achieve high sorting accuracy for products of varying shapes and orientations.
• To validate the system's performance in an industrial context.

Main Methods:

• A vision system was developed for real-time image matching against a stored base image.
• The vision system employs a three-stage process: image binarization, edge detection, and image matching.
• The system operates on a conveyor belt at a speed of 5 cm/s, with sorting decisions made in real-time.

Main Results:

• The developed system successfully sorts products with complex shapes and high efficiency.
• The system demonstrates the ability to detect and sort products regardless of their position or orientation.
• Testing with three different products showed a maximum detection error not exceeding 2%.

Conclusions:

• The automated vision sorting system is effective for industrial applications requiring high-speed, accurate sorting.
• The system's real-time analysis and robust image processing enable reliable identification of diverse products.
• The achieved accuracy and efficiency demonstrate the system's viability for enhancing manufacturing processes.