Breast cancer, a significant health concern among women, requires prompt and accurate diagnosis to improve treatment outcomes and survival rates. Traditionally, specialized doctors perform the diagnosis; however, advancements in machine learning algorithms are enabling supportive diagnostic tools. In this study, we employ a hybrid approach combining Convolutional Neural Networks (CNNs), OpenCV, and Random Forest algorithms to classify breast cancer cases as malignant or benign. The dataset, sourced from the University of Wisconsin, includes 357 malignant and 212 benign tumors, with clinically relevant features extracted using feature engineering techniques. OpenCV is utilized for image preprocessing, ensuring standardized input quality for model analysis, particularly in image-based features. Following data normalization and preprocessing, the dataset is divided into training and testing sets. CNNs are applied to image data for in-depth feature extraction, identifying patterns indicative of malignancy.