Abstract :
This paper introduces the nondestructive prediction of the concentration in trace elements of the quail Coturnix coturnix eggs by color image processing. First, the evaluation of zinc, chrome and magnesium elements of quail eggs was done by atomic absorption spectroscopy. Secondly, algorithms for the determination of the color co-occurrence matrix and Haralick attributes extraction were implemented. Then, eggs were classified according to these attributes and their concentration in trace elements. The obtained results showed that quail Coturnix coturnix eggs are grouped in three distinct classes. They have been determined by evaluating the aspect of the eggshell (Haralick attributes) to deduce the concentration in trace elements of the egg. The attributes f5 (homogeneity), f7 (variance of sums), and f12 (correlation information) are respectively strongly correlated to the zinc, chrome and the magnesium concentrations. Eggs with different shells have shown a different composition in trace elements.
Keyword :
Quail Eggs (Coturnix Coturnix), Trace Elements, Digital Color Image Processing.