Research And Application Of Deep Depression And Weak Edge Medical Image Segmentation Algorithms

Medical image segmentation plays an important role in contemporary medical diagnosis.Accurately extracting clinical information from medical images is one of the key issues of computer-aided medical diagnosis systems.There are a large number of deep depressions and weak edges of contour shapes in medical images,which brings difficulties to certain clinical applications.The GVF Snake model is a widely used algorithm in image segmentation,but in the process of deep concave image segmentation,it is difficult for the external force field to enter the deep concave area,which is likely to cause false edges.At the same time,the GVF Snake model is easily affected by the edge gradient of the image,and the protection effect of the weak edge information of the image is not ideal.Aiming at the problem that the convergence curve of the GVF Snake model is difficult to reach the deep depression area and the weak edge protection ability is poor,this paper has done the following research work.In order to enhance the model’s deep sinking convergence ability,this paper proposes an ALGVF Snake model based on adaptive Laplacian.The model decomposes the Laplacian operator and introduces new coefficients that can be adaptively changed according to the image information before the normal and tangent directions.When it is far from the edge of the image,it enhances the normal direction diffusion to improve the sink convergence ability,and when it is near the edge of the image,it prevents bending.Penetrate weak edges.Add the L~∞(Ω)function to the model to enhance the diffusion in the normal direction.Finally,the component normalization method is used instead of vector normalization to improve the efficiency of curve convergence.Experimental results show that the model can accurately converge to the deep concave area of the image,and the weak edge protection capability has also been greatly improved.Compared with the existing improved model,it has a better segmentation effect.In order to further improve the segmentation effect of the weak edge image,this paper improves the calculation method of the edge graph of the GVF Snake model,and proposes an ECGVF Snake model based on the enhanced Canny operator.Aiming at the problem of the traditional Canny operator losing weak edges,the gradient amplitudes in the two directions of 45° and 135° are increased,and gradient compensation coefficients are added in the horizontal and vertical directions.Experimental results show that using the enhanced Canny operator to calculate the GVF edge map can effectively enhance the strength of the weak edges of the image,so that the model can better protect the weak edge information of the segmented image and improve the accuracy of the image segmentation result.Finally,based on the improvement of the above model,a medical image segmentation system is designed.System design and implementation are carried out from image reading and display,preprocessing,image segmentation based on this algorithm,post processing and other modules.