Characterisation of honey using high-frequency ohmic heating based on image segmentation

Abstract

In the field of computer vision, image segmentation using a clustering approach was employed. This non-destructive method was applied to process ohmic heating in honey, aiming to achieve an efficient and time-saving mass production process. The K-means clustering algorithm converted RGB color data to Lab color space for effective segmentation. The validation of outcomes was conducted through the evolution of RMSE values and regression analysis for each frequency. Notably, at a precision frequency of 1 kHz, the results were as follows: RMSE Red 1.4902, RMSE Green 0.7017, RMSE Blue 0.3328, Regression Red 0.0792, Regression Green 0.5782, Regression Blue 0.202, and heat penetration regression 0.658. This proposed method was benchmarked against the conventional heat penetration analysis in ohmic heating.

Keywords

• heat penetration
• k-means clustering
• ohmic heating
• segmentation