Real-Time Quality Control of Heat Sealed Bottles Using Thermal Images and Artificial Neural Network

Area of Science:

• Industrial Engineering
• Artificial Intelligence
• Computer Vision

Background:

• Effective quality control for heat-sealed bottles is crucial for waste reduction and public health, especially for products like pesticides.
• Traditional inspection methods may be invasive, destructive, or lack automation, limiting scalability and efficiency.
• Assessing seal integrity non-invasively is challenging due to the inaccessibility of seals from the bottle exterior.

Purpose of the Study:

• To design and evaluate an automated, non-invasive, and non-destructive quality control system for heat-sealed bottles.
• To assess the effectiveness of an artificial neural network (ANN) combined with computer vision for seal integrity inspection.
• To demonstrate the system's applicability in industrial settings for pesticide bottle quality control.

Main Methods:

• Development of an automated quality control system utilizing infrared (IR) imaging via a thermal camera.
• Application of computer vision techniques to process IR images of bottle seals.
• Implementation of an artificial neural network (ANN) for seal quality assessment based on image analysis.

Main Results:

• The automated system achieved high accuracy in identifying defective seals, with a precision of 98.6% and a recall of 100%.
• The non-invasive and non-destructive approach using IR imaging proved effective for seal quality inspection.
• The system is designed for seamless integration with existing industrial conveyor belts, enabling large-scale application.

Conclusions:

• The developed AI-powered system offers an effective solution for quality control of heat-sealed bottles, particularly for hazardous contents like pesticides.
• The automated, non-destructive inspection method enhances efficiency, reduces waste, and ensures product safety.
• This technology is scalable and adaptable for widespread use in high-volume industrial packaging lines.