Bone Fracture Detection System using CNN Algorithm

Abstract

Identification of faults through computer-based techniques is a growing trend these days in all fields. A highly responsive system is characterized by two key features of quick detection and being highly accurate through leverage of modern techniques and efficient utilization of resources. Break in a bone or bone fracture is the result of excess external force beyond the threshold of what the bone can withstand. Canny Edge detection is an image processing methodology to detect the bone fracture through efficient use of automated fracture detection and overwhelms the noise removal problem. In today’s world there are several methodologies available for edge detection like Sobel, Canny, Log, Prewitt, and Robert. However, these techniques are plagued with key shortcomings like a lack of capability to perform multiresolution analysis that result in inability to detect minor details during analysis. The other key shortcoming of the techniques is that though they work fine with high resolution and high-quality images, but can’t work as well with noisy images due to their inherent lack of ability to distinguish between edges and noise components [4]. The method being proposed overcomes over these problems using CNN algorithm. The results from the simulations done reveal that the proposed method is much more efficient mechanism to perform edge detection at aggregate scales. The proposed method has also proved to be resilient enough to extract the necessary information and do the processing needed on key portions of the images and handle noise in a much better manner than the currently available edge detectors