<![CDATA[The egg hatching process is a crucial stage in chicken breeding, as the quality of the resulting embryos will greatly determine the productivity and success of the cultivation. However, monitoring the internal condition of eggs during incubation is still limited and is generally done manually using the candling method. Therefore, a non-invasive approach is needed to support the process of monitoring embryo development. This study aims to develop a chicken embryo classification system based on in-ovo images using a Convolutional Neural Network (CNN). This system utilizes internal egg images taken with a camera integrated in the incubator, so that the condition of the eggs can be automatically classified into categories of fertile, infertile, and developmental failure. The research stages include collecting in-ovo image data, preprocessing in the form of removing the background using the rembg library, and resizing the images to 224 × 224 pixels. The dataset was then divided into training data and test data with a ratio of 90%:10%. The CNN model was built with several layers, namely convolution, max pooling, dropout, flattening, and fully connected, then trained for 28 epochs with a batch size of 32. The results showed that the developed CNN model was able to achieve an accuracy of 90.5% in the 24th epoch and experienced a decrease in the 26th to 28th epochs. Testing also used images from different incubators than those used from the dataset showed that the model could classify egg conditions well.]]>
