This study delves into the complex issue posed by facial makeup, which has the potential to significantly alter the appearance of individuals, posing a challenge to automated face recognition systems, as well as age and beauty estimation methods. A model solution aimed at automatically detecting makeup in facial images to improve the accuracy of recognition systems was proposed in this work. The approach revolves around utilizing a sophisticated model that harnesses a feature vector encapsulating crucial aspects of facial attributes including shape, texture, and color. Employing an advanced Convolutional Neural Network (CNN) architecture, the model detects the presence of makeup by analyzing key facial landmarks such as eye distance, nose width, eye socket depth, cheekbones, jawline, and chin. Experiments were performed on a dataset consisting of 200 facial images to assess the effectiveness of the proposed method. The model achieved a validation accuracy of 100%, demonstrating its robustness in makeup face detection. Notably, the computational runtime for the validation process was 1 minute and 40 seconds, underscoring the efficiency of the proposed approach. Moreover, an innovative adaptive pre-processing strategy that capitalizes on makeup information to enhance the performance of facial recognition systems was developed. This strategy aims to optimize the recognition process by leveraging insights gained from makeup detection. By integrating this adaptive pre-processing step, further advancements in the accuracy and reliability of facial recognition technology, particularly in scenarios where makeup may confound traditional recognition methods, are envisioned.