A Study Of Medical Image Segmentation Based On Wavelet Analysis And Gradient Vector Flow

Medical image segmentation technology plays an important role in digital medical image processing. Clinical medical images are often susceptible to various conditions.Images are affected by noise and complex shapes. When images are extracted, they usually produced pseudo-edge and inaccurate results.Therefore, many researchers have improved existing methods, The effects of deformation model have more far-reaching. It uses energy minimization process iterated to edge of the image, this method is widely applied to image segmentation and so on. The traditional deformation model includes the snake, balloon models and so on, Recently, gradient vector flow field model (Gradient Vector Field, GVF) is widely better used in the field of image processing.GVF deformable model has been greatly improved on initialization and capture range. But for complex medical images, the initialization position has more stringent requirements. It is expensive in computation and a lot of time due to noise. In this paper, a multi-resolution wavelet analysis approach is proposed to solve the problems. Through wavelet multi-scale transform, it extracts the image edge and improves the anti-noise performance. The combination of wavelet analysis method and GVF deformable model have better segmentation on medical images.The improved method has been applied to clinical images. It is accurate and rapid convergence to the edge of medical images, Experiments have demonstrated that the new method is effective and has better robustness and higher accuracy.