Automatic detection of fracture in femur bones using image processing

The femur is the enlarged and the vigorous bone in the human body, ranging from the hip to the knee. This bone is responsible for the creation of Red Blood Cell in the body. Since this bone is a major part of the body, a method is proposed through this paper to visualize and classify deformities for locating fractures in the femur through image processing techniques. The input image is preprocessed to highlight the domain of interest. In the process, the foreground which is the major domain of interest is figured out by suppressing the background details. The mathematical morphological techniques are used for these operations. With the help of basic morphological operations, the foreground is highlighted and edge detection is used to highlight the objects in the foreground. The processed image is classified using the support vector machine (SVM) to distinguish fractured and unfractured sides of the bone.