The edge detection of brain tumor

The edge detection plays a critical role in medical application, Edges characterize boundaries are fundamental problem in image processing. The paper illustrates the way by canny algorithm detecting the weak edge of brain,then labeling all the 8-connected edge with a different number and classifying with the that edge, for the size of all the 8-connected edge circumference being different, So we can plot the histogram according to the size of the edge, At last, we can detect the sole weak edge by the histogram segmentation. This way can help us get the deformable edge such as brain tumor.

[1]  Richard W. Hall Optimally Small Operator Supports for Fully Parallel Thinning Algorithms , 1993, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  John F. Canny,et al.  A Computational Approach to Edge Detection , 1986, IEEE Transactions on Pattern Analysis and Machine Intelligence.