Enhancing Facial Expression Recognition Accuracy Through Haar Cascade-Based Feature Extraction

Facial Expression Recognition (FER) has a significant interest because of its increased applications in Human-Computer Interaction (HCI), like video interactions, emotion analysis, image indexing, and retrieval and many more. In the discipline of computer vision, pattern recognition, and artificial intelligence FER has become an extremely active research topic. A crucial component of FER systems is feature extraction, which involves identifying unique features from facial images and representing them in a quantized form. This research work utilizes a robust feature extraction method namely the Haar Cascade Algorithm. This work is motivated by the drawbacks in the existing feature extracting techniques which decreases the accuracy in predicting facial expressions. The main objective of this work is to demonstrate the impact of Haar Cascade based feature extraction technique in the accuracy of predicting emotions. Previous studies are reviewed to evaluate the effectiveness of existing feature extraction methods. Based on this analysis, the most suitable approach is identified to improve accuracy in recognizing the seven basic facial expressions: happiness, sadness, anger, surprise, disgust, fear, and neutral.