This study proposes an automated system for assessing the freshness of chicken eggs through ovoscopy image analysis using Convolutional Neural Networks. Accurate freshness evaluation is essential for food safety and public health, particularly in regions where eggs are stored at ambient temperature, and for optimizing supply chains to reduce food waste. Ovoscopic images were acquired daily from eggs stored under ambient conditions over a 30-day period. The dataset was categorized into 30 classes, each representing a specific day of storage. The ResNet-18 architecture was implemented using transfer learning, with its fully connected layers adapted for the freshness classification task. Model performance was evaluated using classification accuracy, achieving a result of 80.36%. While the results indicate promising performance, especially at the extremes of the storage period, the model showed difficulty in distinguishing between consecutive days due to subtle visual changes. These findings suggest potential for practical implementation, although further validation with larger and more diverse datasets is needed to confirm generalizability and robustness.