Contour Detection of Human Knee

Medical images (X-ray, CT, MR or PET) of human organs are widely used in the everyday clinical praxis. The paper presents methods which allows the user to analyze, model and adjust the total knee contour detection procedure during a preoperative planning which eliminating the need of using expensive and sophisticated commercial software tools, which procedures are not familiar to the natural user, the surgeon. It has based on a observation which is affects the global saliency of the contour from the local context of a edges. Here we have propose the scheme consists of first find the edge in which response all points in an image using gradient computation and in the second one adapted the edge response at a point by the response in its border. Here we also apply the canny edge detection technique to present the results of implementing the scheme followed by a mask operation for the border manipulation. We have examined on a large set of images which is tested successfully. This scheme is computationally and qualitatively favorably in comparison with another human knee contour detector technique which is also based several influence. The proposed scheme can provides a little expenditure preprocessing step for high level tasks such shape based human knee contour detection.