The Freshness Assessment of Eggs Based on Thermographic Image Processing Using Blob Detection Method

Eggs are food ingredients that people often consume daily because they are a good source of protein for the body. However, long storage time can cause eggs to be unfit for consumption; therefore, rotten eggs are often found in the market. Identifying eggs is usually done by smelling their aroma and analyzing the light that passes through them, which is subjective. This study develops a system to assess the freshness of chicken eggs using a thermal camera through thermography analysis. As the storage time increases, the air cavity in the egg gets bigger. The cavity will inhibit the temperature of the liquid, leading to the thermal camera. A Raspberry Pi single-board computer processes the thermography image produced by the AMG8833 thermal camera. This thermography image processing includes interpolation, hue saturation value (HSV) transformation, morphological operation, contour detection, and blob detection. The liquid's blob area and the egg's total blob are used as freshness parameters, including fresh, old, and rotten. Principal Component Analysis (PCA) is employed to visualize the classification of egg freshness observed in about a month. The experimental results show that each egg's freshness can be separated in the PCA feature space. In addition, the relative ratio of the liquid area and the total area has a linear relationship with the freshness of eggs. In a test involving 50 eggs from the market and chicken farm, this system can distinguish its freshness with an accuracy of 94.9%. In future work, testing under real conditions will be required to obtain robust equipment. This system can assess egg freshness noninvasively, quickly, and objectively.