Online Machine Vision-Based Modeling during Cantaloupe Microwave Drying Utilizing Extreme Learning Machine and...

Area of Science:

• Food Science and Technology
• Process Engineering
• Machine Vision Applications

Background:

• Online monitoring of microwave drying is difficult due to metal component interference.
• Existing methods lack real-time material characteristic tracking.

Purpose of the Study:

• To develop a non-invasive microwave drying monitoring system using machine vision.
• To model and predict material moisture content during drying using shrinkage data.
• To compare the predictive performance of artificial neural networks (ANN) and extreme learning machines (ELM).

Main Methods:

• A microwave drying system integrated with online machine vision for real-time image, weight, and temperature acquisition.
• Development of an image-processing algorithm to quantify material shrinkage.
• Application of artificial neural network (ANN) and extreme learning machine (ELM) for moisture content prediction based on shrinkage.
• Constant-temperature microwave drying experiments were conducted.

Main Results:

• The developed system and image-processing algorithm effectively monitored microwave drying in real time.
• Both ANN and ELM models successfully predicted moisture content based on material shrinkage.
• Extreme learning machine (ELM) demonstrated superior predictive accuracy and learning efficiency over artificial neural network (ANN).

Conclusions:

• The machine vision-based system offers a viable solution for real-time microwave drying monitoring.
• ELM provides a more efficient and accurate approach for predicting moisture content compared to ANN in this application.
• This technology enables enhanced process control and optimization in microwave drying.