Coronary Angiography Image Edge Detection Based on Information Measure

A new edge detection method is presented based on the information measure, and the new method is applied to the coronary angiogram. Firstly, the blood vessel path points are found by using the square template, and the quasi-center curve of the vessel segment is obtained by using the minimum gray scale criteria and the smoothing technical. Then the orientation information measure concept of an image is introduced, and the l values of the information measures of the l points in the same row or the same column are calculated according to the quasi center curve. Finally, the point with the maximum information measure is defined to be the edge point and the edge curve is acquired. The experimental results show that the new method can effectively distinguish the noise points and identify the edge points with the better accuracy. This method has a good immunity in noise. The experimental result is quite satisfactory.