Ensuring the security of systems utilizing face authentication for accessing services and information has become crucial. While past research has applied machine learning and deep learning techniques for spoof detection, challenges persist in image preprocessing and feature extraction, which are essential for enhancing the effectiveness of face anti-spoofing models. This study delves into diverse aspects of dataset preprocessing to refine face spoof detection. The research investigates the impact of background removal using the Viola-Jones algorithm. It explores different color spaces (YuV, CIE LAB, YCbCr, and HSV) to identify the most suitable options for generating robust face-spoof detection features. The proposed approach involves employing a convolutional neural network (CNN) for feature extraction and support vector machine (SVM) to classify real and spoof images, aiming to enhance the overall performance of the detection method.