Bone fractures are common injuries that need timely and accurate diagnosis. Conventional diagnostic procedures, such as X-ray imaging, are critical in finding fractures, but they usually depend on manual interpretation, which can lead to human errors and inefficiency. To solve these issues, our study intends to create an automated method for detecting bone fractures us-ing machine learning and image processing. This method consists of several phases, including feature extraction, edge detection, pre-processing, as well as machine learning categorization. In addition to identifying the existence of a broken bone in X-ray pictures, this system will la-bel the position of several fracture types within the image. To achieve high accuracy, a modern object recognition algorithm is trained using a collection of X-ray pictures. By uploading their x-ray photos to the platform, customers will be able to examine and detect fractures remotely thanks to a website that has been built. The model achieved an accuracy of 76% and processed images at an average speed of 30 ms per image.