Brain Tumor Detection in Medical Images Using HOG Feature Extraction and Machine Learning Algorithms

Abstract

Brain tumor detection is a critical task in medical image analysis. Early and accurate diagnosis helps improve patient survival rates. Medical imaging techniques such as MRI are commonly used. Manual analysis of MRI images is time-consuming. It also depends heavily on expert radiologists. Automated systems can assist in diagnosis. This project proposes a machine learning-based approach. Histogram of Oriented Gradients (HOG) is used for feature extraction. HOG captures important texture and shape features. Extracted features are used for classification. Machine learning algorithms analyse these features. The system classifies images as tumor or non-tumor. Preprocessing improves image quality. Noise reduction enhances feature extraction. The approach reduces human effort. It improves diagnostic accuracy. The system is cost-effective. It supports faster decision-making. The model is scalable. The results show promising performance.