Cancer remains a prominent global cause of mortality, mandating early detection for effective treatment and reduced fatality rates. Accurate lesion classification is pivotal in controlling cancer-related mortality. However, this process is inherently complex, relying heavily on microscopic image analysis to identify malignancies. Deep neural networks, particularly Convolutional Neural Networks (CNNs), have proven invaluable in various image-processing tasks, including segmentation, registration, and classification. This paper proposes an innovative approach that harnesses the power of CNNs, specifically DenseNet-121, by extracting Texture features like Histogram of Oriented Gradients (HOG) and Local Binary Patterns (LBP) to efficiently analyze medical data. Texture Features are extracted based on variations in pixel intensities within the image. These features are passed to SoftMax function for classification of breast cancer into benign and malignant, which is crucial for early and effective treatment, ultimately reducing cancer-related mortality rates.