Aiming at the complex background of coronary angiograms, weak contrast between the coronary arteries and the background, a new segmentation method based on transition region extraction is proposed. Firstly, the coronary arteries are extracted by using the local complexity method based on Top-hat. Then the coronary arteries are extracted again by using the local complexity method based on Gaussian filter. Finally, the segmentation image is obtained by fusing two extracted coronary arteries images. The experiments indicate that the proposed method has better performance on the small vessels extraction and background elimination. In addition, the method is valuable for diagnosis and the quantitative analysis of vessels.