Weight estimation, shape index determination and shell contamination detection of hen egg using machine vision methods

This study introduces a method based on machine vision technology for weight estimation, shape index determination, and egg contamination detection. Primarily, different characteristics relevant to egg quality such as weight, length, wide, shape index and contamination statue of 76 eggs were recorded by experienced appraisers. At the same time, using a digital camera several photos were taken from each egg regarding fixed imaging distance and unique illumination. Then, the features relevant to egg weight, egg size and egg shell contamination were extracted from digital images using image processing tools (IPT) of MATLAB software. Four artificial neural networks were designed for egg contamination classification, egg weight, and egg length and egg wide estimation respectively using MATLAB software. The first neural network was trained to detect egg contamination without error, and in the test phase, the eggs contamination was recognized with accuracy of 95% by neural network. Next neural networks were trained to estimate the egg weight, egg length and egg wide with accuracy of 98.55, 99.4 and 98.49 % respectively. In the test phase, the correlations between egg weight, egg length and egg wide with those which estimated by artificial neural networks were equal to 96.6, 97.31 and 98% respectively (P