CIMOR: An automatic segmentation to extract bone tissue in hand x-ray images

This paper presents a new method to segment bone tissue in an X-ray images. The proposed method is completely automatic. First of all, it enhances the contrast of the grayscale image using contrast-limited adaptive histogram equalization. Afterwards, the image intensity values are further modified by removing the background and soft tissue. Then a series of morphological operators, a opening-by-reconstruction followed by a closing-by-reconstruction, are applied to enhance the image by defining foreground objects and clean up the image. Finally holes are filled to preserve the bone structure.