Abstract Manual external egg quality testing often faces various problems, one of which is the low level of accuracy. Testing processes that rely on human ability to identify eggshell defects can produce inconsistent results. So more efficient and accurate quality testing method is needed, namely imaging technology. Using digital imaging to evaluate egg quality offers advantages, including the ability to automatically and consistently detect shell defects. In applying digital image classification, good training data is needed to get accurate classification results, so the aim of this research is to extract chicken egg color features as training data to optimize the classification system. The method used on classifying chicken eggs based on shell color is feature extraction based on color in HSV (Hue Saturation Value) and RGB (Red Green Blue). Research results obtained 95 image data of chicken eggs consisting of 61 Grade A eggs with average RGB value of 75.39 and average HSV value of 0.36, 28 Grade B eggs with average RGB value of 87.92 and average HSV value is 0.35, and 6 Grade C eggs have average RGB value of 92.36 and average HSV value of 0.33, with color extraction in HSV and RGB spaces.
Read full abstract