Accurate recognition of facial expressions in images is vital as it provides deeper insights into emotional states, enhances human-machine interactions, and has significant practical applications in areas such as mental health assessment and customer feedback analysis. This study aims to develop advanced machine-learning models for the precise recognition of various facial expressions in images. Traditional methods of gauging human emotions often fall short due to their high cost and inaccuracy, highlighting the need for more effective solutions. To address this, we implemented two classification algorithms, Support Vector Machine (SVM) and Convolutional Neural Network (CNN), and evaluated them on two image databases (JAFFE and CK+). These databases contain images depicting seven emotions: “happy”, “sad”, “disgust”, “surprise”, “neutral”, “angry”, and “fear”. For the SVM model, we utilized a geometry-based feature extraction technique, while for the CNN model, we introduced a novel feature extraction method. The results demonstrated enhanced performance in recognizing facial expressions, thereby underscoring the potential of these models to improve human-computer interaction and other practical applications significantly.