Abstract: In order to clarify the texture thickness and tissue distribution of the X-ray image, enhance the presentation of body structure information, and reduce the wrong judgment of fuzzy image on doctors’  diagnosis results, we proposed an adaptive enhancement algorithm of digital X-ray image based on Markov random field model. Firstly, the algorithm counted the pixels with the same brightness in the whole range of X-ray image, and the histogram equalization method was used to transform the original image into gray level distribution image  to eliminate light interference. Secondly, we analyzed the organization attributes, extracted the texture features of X-ray image through gray level co-occurrence matrix, and obtained the gray level information of image texture thickness and layout structure. Finally, the average brightness was calculated by the mapping function and logarithmic function, the Markov random field model was used to adjust the brightness of the image, supplement the brightness of small parts of the texture, then the smooth image was divided by the random field function, and the  secondary reconstruction was adopted to ensure the balance of image sharpening and enhancement effect. The simulation results show that the proposed algorithm can improve the internal information clarity of the image.

Key words: Markov random field model, digital X-ray image, image adaptive enhancement, image feature extraction, image preprocessing