Research On Body Cell Image Segmentation Technology Based On Multi-information And Multi-scale

Medical image segmentation is the key technology of medical image processing,and it is also the basis of image analysis and recognition and the application of medical image.The technology of image segmentation is more and more important in clinical diagnosis and auxiliary treatment.It is a difficult and arduous task to automatically segment the object of interest from medical images,In addition to the complex diversity of the medical image itself,there will be some noise in the medical image during the imaging process.In addition,the result of the segmentation algorithm is also Will be affected by partial field shift effect,local body effect,gray unevenness,artifact and other factors.The traditional segmentation method is difficult to meet the needs of medical image segmentation,so the medical image segmentation method for in-depth study is very necessary.Wavelet transform has good local detection ability and multiresolution analysis in time domain and frequency domain.This is the theoretical basis for applying wavelet transform to cell image segmentation.The traditional edge detection operator is generally gray target edge discontinuity using this characteristic,by calculating the gradient of the local extremum pixels,the pixel is connected with the target edge,but is easily disturbed by noise information.In this paper,an improved B-spline edge detection algorithm with adaptive threshold is proposed,In order to obtain accurate edge information and remove the noise,the quadratic B-spline function is used as the wavelet function,and the local modulus maxima is calculated by using the porous algorithm.Then,the adaptive threshold is automatically raised according to the characteristics of edge and noise.The separation of noise and edge,the separation of strong edge and weak edge,and the multi-scale matching and fusion strategy.Finally,the edge of the cell image is obtained,and the effectiveness of the algorithm is verified by experimental analysis.Medical image processing to extract cells in the use of watershed method,prone to over-segmentation of the phenomenon and the interference is very sensitive to noise,in order to solve this shortcoming,this paper proposes a new method of image segmentation based on wavelet transform and watershed.In this algorithm,the wavelet transform is used to decompose the image,and the appropriate wavelet base and the improved denoising threshold function are selected to denoise the image,then,the reconstructed cell images of denoised wavelet are transformed into the watersheds generated by mathematical morphological distance transformation and gray scale reconstruction,and finally the segmentation results are obtained.The experimental results show that the algorithm can accurately and accurately extract the cells and realize the automatic segmentation of the adhesion cells,and has good robustness and universality.