The Medical Image Segmentation Based On The Wavelet Transform

Image segmentation is an important image analysis technology which has been widely used practically and paid close attention to.Meanwhile,image segmentation is a key step from image process to image analysis and plays an important role in image engineering.It is the base of target express and it has an important effect on characters measuration,at the same time which makes it possible to carry on deeply image analysis and image understanding.Wavelet Transform is a time-frequency 2-field analytical tool and of good local characters in both fields.Because of this it can take effective information from signals.Wavelet Transform deals functions and signals with multiresolution analysis,especially,when we face with unsteady signals wavelet transform changes its resolution in accordance with the change of signals which are waiting to be analyzed,this is so called multiresolution analysis,so wavelet transform is more useful and can meet the practical needs.In this paper it first gives a brief summary of the existing basic image segmentation methods, introduces the traditional method of image segmentation:threshold segmentation, regional growth technology, edge detection, image segmentation based on the fractal and image segmentation based on neural network technology. For complex medical CT, MRI images, combined with the merits of wavelet transform to deal with images, we focused on introducing wavelet transform theory and the properties of wavelet transform, and then filter the analytical 256×256 image to eliminate noise, in another word called preprocess-image.In this thesis it represents some research of the image segmentation based on histogram wavelet transform, calculates the wavelet transform of the image statistics histogram, and by the coarseness and fine seeking the segmentation threshold and obtained better segmentation results. At the same time, segments the histogram combined with the smallest error means, accesses the comparison effects graph of the traditional histogram segmentation and wavelet coefficient histogram segmentation of the image. In addition, this paper represents the edge detection method based on wavelet transform. One method of retaining the high-frequency component of the image is decomposing the image by wavelet transform, to get the approximate low-frequency components of the images and horizontal, vertical, diagonal part of the high-frequency components, and then using wavelet reconstructed algorithm to reconstruct the high-frequency part of the image and than extracts the edge information.In the last part of the thesis it describes the evaluation criteria of segmentation, introduces mathematical morphology to determine the segmentation evaluation parameters of our image segmentation and determine the quality of image segmentation.