Abstract: Computer Aided Diagnosis (CAD) combining engineering approach and medical approach is popular in medical works. Bone fracture detection system which is medical image processing is one important system in CAD field. The purpose of this work is that Lower Leg Bone (Tibia) in x-ray image is used to detect fracture or not by computer vision system. Bone fracture is a common problem which is mostly encountered in recent year. Using image processing tools, bone injury diagnosis is developed to be more accurate and reliable. Generally, fracture detection systems use image processing tools in matlab based on mathematical calculation. Automatic fracture detection system can effectively aid to doctors and radiologists to overcome heavy workload. The proposed system was implemented by preprocessing, feature extraction and classification. Sharpening is used for preprocessing, Harris corner detection is used for corner feature and SVM is used for classification. According to the result, the algorithm can correctly classify normal bone and broken bone with good accurate and quick rate with efficient features and single classifier.